Meeting session control based on attention judgment

Abstract

A system and method are provided for meeting session control based on attention determination. The system receives a plurality of images of a plurality of attendees associated with a plurality of meeting sessions. The system detects one or more activities performed by each of the plurality of attendees during the corresponding meeting session over a period of time. The system calculates an attention score for each of the plurality of attendees over the corresponding period of time based on the detected one or more activities associated with the corresponding attendee. The attention score is indicative of each attendee's level of attention in the corresponding meeting session. The system further trains a machine learning (ML) model for each of the plurality of attendees based on the calculated attention score for each of the plurality of attendees and a meeting category of the corresponding meeting session.

Description

[Technical Field]

[0001] Cross-reference to related applications / incorporation by reference This application claims the benefit of priority to U.S. Patent Application No. 17 / 671,495, filed with the United States Patent and Trademark Office on February 14, 2022. Each of the above applications is incorporated herein by reference in its entirety.

[0002] TECHNICAL FIELD Various embodiments of the present disclosure relate to meeting sessions, and more particularly, to systems and methods for meeting session control based on attention determination. [Background technology]

[0003] Advances in the field of machine learning have led to multiple applications in various areas of life. With the outbreak of certain pandemics, almost all educational institutions, offices, and meeting spaces have been closed or significantly affected. Therefore, multiple students and / or employees have been forced to learn and / or attend meetings remotely via the internet. However, remote learning and remote meetings can pose different challenges, including (but not limited to) the difficulty of maintaining each student's or employee's attention level. Some solutions focus on content delivery rather than participant attention. Therefore, there is a need for a high-impact interactive system that takes into account the attention of attendees present in a meeting session and further enhances the attendees' overall learning process. Summary of the Invention [Problem to be solved by the invention]

[0004] The limitations and disadvantages of conventional approaches will become apparent to those skilled in the art by comparing the described system with certain aspects of the present disclosure illustrated in the remainder of this application and with reference to the drawings. [Means for solving the problem]

[0005] A system and method for attention-determining-based meeting session control is provided substantially as shown and / or described in connection with at least one of the drawings and more fully set forth in the claims.

[0006] These and other features and advantages of the present disclosure will become apparent from a consideration of the following detailed description of the disclosure when taken in conjunction with the accompanying drawings, in which like reference characters refer to like elements throughout. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a block diagram illustrating an exemplary network environment for attention-determination-based meeting session control, according to an embodiment of the present disclosure. [Figure 2] 1 is a block diagram illustrating an exemplary system for meeting session control based on attention determination, according to an embodiment of the present disclosure. [Figure 3] FIG. 1 illustrates an example operation for training an ML model for each of multiple attendees associated with multiple meeting sessions, according to an embodiment of the present disclosure. [Figure 4] 4 illustrates exemplary operations for attention determination and meeting session control based on application of the ML model for a particular attendee trained in FIG. 3, according to an embodiment of the present disclosure. [Figure 5] FIG. 1 illustrates an exemplary scenario for generating a simulated view of multiple attendees associated with a meeting session, according to an embodiment of the present disclosure. [Figure 6] FIG. 1 illustrates an exemplary user interface for rendering dashboard information, according to an embodiment of the present disclosure. [Figure 7] 10 is a flowchart illustrating an example operation of training an ML model for attention-determination-based meeting session control according to an embodiment of the present disclosure. [Figure 8] 1 is a flowchart illustrating exemplary operations for meeting session control based on attention determination and application of ML models for particular attendees, according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0008] The disclosed system and method for meeting session control based on attention determination may include the following embodiments. When executed on a system, such a method may be configured to train a machine learning (ML) model for attendees (such as students or employees) based on the attendees' attention levels determined in different meeting sessions (such as classes or conferences in offline or online mode). The system may also generate one or more recommendations in real time for increasing the attendees' attention levels based on application of the trained ML model associated with the corresponding attendees. The system may receive multiple images of multiple attendees associated with multiple meeting sessions to train the ML model. The multiple images may be received from different imaging devices (such as cameras) located during the meeting sessions. One or more of the multiple meeting sessions may have different meeting categories. The meeting category may be, for example, the type of meeting (i.e., class / offline-based or online-based), the number of attendees in the meeting session, the duration of the meeting session, the average age of the attendees, the topic of the meeting session, the content presented during the meeting session, or the experience of the educator in the meeting session. The system may further detect, based on the received plurality of images, one or more activities (e.g., raising a hand, reading, writing, typing, talking, standing, asking / answering questions, etc.) performed by each of the plurality of attendees during the corresponding meeting session. The one or more activities may be detected over a period of time (e.g., over several minutes / hours in a particular meeting session). The system may further calculate attention scores for each of the plurality of attendees over the corresponding period of time based on the one or more detected activities associated with the corresponding attendee. The attention scores may indicate the attention level of each attendee in the corresponding meeting session.The system may be further configured to train a machine learning (ML) model for each of the multiple attendees based on the calculated attention score for each of the multiple attendees and the meeting category of the corresponding meeting session. The disclosed system may be capable of training multiple ML models for the multiple attendees according to the attention scores calculated for corresponding meeting sessions of different meeting categories. Thus, the disclosed system may be configured to generate and train a personalized machine learning (ML) model for each attendee. The disclosed system may train the ML model based on attention levels determined during multiple meeting sessions of the same or different meeting categories (i.e., attention levels calculated based on one or more activities performed by the attendee). The generated ML model may also be further trained based on multiple other factors, such as the attendee's facial expressions in different meeting sessions, experience information related to the educator, environmental information (such as lighting conditions in the meeting session), and the attendee's preferences / interests. Thus, a personalized ML model may be trained for a particular attendee based on attention scores calculated (or tracked) for the attendee in different meeting sessions having different meeting categories and different associated other factors. In other words, the trained ML model for attendees can represent historical data and fluctuations / patterns related to attention scores that can be determined for attendees of different categories of past meeting sessions and meeting sessions with other related factors. Thus, the disclosed system can propose a characteristic learning model for each student (i.e., attendee) by processing all historical data (related to attention scores) generated from a series of class activities over a period of time. This data is useful for conducting cumulative analysis of each student, thus identifying each student's strengths and weaknesses. The disclosed system can further consider a student's tendency toward certain subjects and their responsiveness to teaching methods.

[0009] Based on the generation of the ML model, the disclosed system can further apply the trained ML model to attendee images or calculated attention scores (i.e., associated with the current or real-time meeting session) to further output or control a set of recommendations for attendees, educators, or content to be presented in the meeting session. One or more recommendations, if followed, can increase the attendee's attention level in the current or upcoming meeting session. Based on the real-time analysis of the attention level and the output recommendations, the corresponding meeting session can be controlled to increase the attendee's attention level.

[0010] The disclosed system can be an artificial intelligence (AI)-based intelligent system that can aim to engage students (i.e., attendees) in the learning process by enabling educators in a meeting session to effectively transfer / transmit knowledge to multiple attendees (without disrupting the essence of regular classes). The disclosed system can also be implemented as a smart teaching mechanism that can monitor the attention level of each attendee (using a camera and depth sensor) and aim to further improve the overall learning process in real time based on the monitored attention level of each attendee and output recommendations for the attendee, educator, or presented content. The disclosed system can also evaluate the quality of the lecture and instructor of a meeting session using a real-time analysis of attendee attention levels, utilizing cumulative data (i.e., an ML model trained based on determined attention scores of different meeting sessions). Furthermore, the disclosed system can provide detailed time-varied individual reports for each attendee in addition to cumulative group analysis. Thus, the disclosed system can focus on each individual attendee to improve their learning process. Furthermore, the disclosed system can generate various reports and statistics based on multiple parameters such as subjects and grades (etc.) to perform comparative analysis of each attendee with other attendees present in the meeting session. Furthermore, the disclosed system can also detect illegal activities (such as cheating during an exam) that may be committed by attendees or educators during the meeting session. The disclosed system can also have a prediction mechanism that can adapt to each student's learning pattern.

[0011] 1 is a block diagram illustrating an exemplary network environment for attention-determination-based meeting session control, according to an embodiment of the present disclosure. Shown in FIG. 1 is a network environment 100. The network environment 100 may include a system 102, multiple image capture devices 104, multiple machine learning (ML) models 106, an audio capture device 108, a server 110, and a communication network 112. Also shown in FIG. 1 are multiple images 114 of multiple attendees 116.

[0012] The system 102 may include suitable logic, circuitry, interfaces, and / or code that may be configured to receive a plurality of images 114 of a plurality of attendees 116 associated with a plurality of meeting sessions in which the corresponding attendees are present (e.g., meeting session 118 shown in FIG. 1 ). The system 102 may be configured to calculate an attention score for each of the plurality of attendees based on the received plurality of images 114, and further train a plurality of ML models 106 for the plurality of attendees 116 based on the calculated attention score for each of the plurality of attendees 116. Examples of the system 102 may include, but are not limited to, a computing device such as an educational engine, a computer workstation, a mainframe machine, a server, a smartphone, a mobile phone, a personal computer with or without a graphics processing unit (GPU), an imaging device with processing capabilities, and / or a consumer electronics (CE) device. In some embodiments, the system 102 may be implemented as a plug-in that can be added to existing video teleconferencing software programs and applications.

[0013] Each of the multiple image capture devices 104 may include suitable logic, circuitry, and interfaces that can be configured to capture multiple images 114 of multiple attendees 116. Each of the multiple image capture devices 104 may be further configured to transmit the captured multiple images 114 to the system 102. Examples of each of the multiple image capture devices 104 may include, but are not limited to, an image sensor, a closed-circuit television (CCTV) camera, a webcam (or web camera), a wide-angle camera, an action camera, a camcorder, a digital camera, a camera phone, a time-of-flight camera (ToF camera), a night vision camera, and / or other image capture devices. In some embodiments, the multiple image capture devices 104 may include a depth sensor that can be configured to capture depth information / multiple depth values ​​of the multiple attendees 116 from a single perspective or multiple perspectives in a corresponding meeting session (e.g., a lecture).

[0014] Each of the multiple machine learning (ML) models 106 can be an untrained classifier / regression / clustering model that needs to be trained to identify relationships between inputs, such as features, in a training dataset and output a recommendation set. Each of the multiple ML models 106 can be defined by hyperparameters, such as the number of weights, a cost function, an input size, and the number of layers. The parameters of each of the multiple ML models 106 can be adjusted to move toward a global minimum of the corresponding ML model's cost function, and the weights can be updated accordingly. Each of the multiple ML models 106 can be trained to output a prediction / classification result for an input set after several epochs of training on feature information in the training dataset. The prediction result can indicate a class label for each input in the input set. In one embodiment, each of the multiple ML models 106 can be trained for a different attendee based on attention scores calculated from past meeting sessions attended by the corresponding attendee, which can have different meeting categories.

[0015] Each of the plurality of ML models 106 may include electronic data that may be implemented, for example, as a software component of an application executable on the system 102. Each of the plurality of ML models 106 may rely on libraries, external scripts, or other logic / instructions for execution by a device, such as a circuit. Each of the plurality of ML models 106 may include code and routines configured to enable a computing device, such as the system 102, to perform one or more operations to output a recommendation set. Additionally or alternatively, each of the plurality of ML models 106 may be implemented using hardware, including a processor, a microprocessor (e.g., that performs or controls the execution of one or more operations), a field programmable gate array (FPGA), or an application-specific integrated circuit (ASIC). Alternatively, in some embodiments, each of the plurality of ML models 106 may be implemented using a combination of hardware and software.

[0016] In one embodiment, each of the plurality of ML models 106 may be implemented as a neural network model, such as a deep learning model. A neural network model may be defined by its hyperparameters and topology / architecture. For example, a neural network model may be a deep neural network-based model that may have a number of nodes (or neurons), activation function(s), number of weights, cost function, regularization function, input size, learning rate, number of layers, etc. Such a model may be referred to as a computational network or a system of nodes (e.g., artificial neurons). In a neural network implementation, the nodes of the neural network model may be arranged in layers as defined by the neural network topology. These layers may include an input layer, one or more hidden layers, and an output layer. Each layer may include one or more nodes (or artificial neurons, e.g., represented by circles). The output of every node in the input layer may be connected to at least one node in the hidden layer(s). Similarly, the input of each hidden layer may be connected to the output of at least one node in another layer of the model. The output of each hidden layer may be connected to the input of at least one node in another layer of the neural network model. The node(s) in the final layer can receive input from at least one hidden layer and output a result. The number of layers and the number of nodes in each layer can be determined from hyperparameters that can be set before, during, or after training of the neural network model based on the training dataset.

[0017] Each node in a neural network model can correspond to a mathematical function (e.g., a sigmoid function or a rectified linear unit) with a set of parameters that can be adjusted during training of the model. The set of parameters can include, for example, weight parameters and regularization parameters. Each node can calculate an output using a mathematical function based on one or more inputs from nodes in other layer(s) (e.g., previous layer(s)) of the neural network model. All or some of the nodes in a neural network model can correspond to the same or different mathematical functions.

[0018] In training a neural network model, one or more parameters of each node can be updated based on whether the output of the final layer for a given input (from the training dataset) matches the correct result based on the neural network model's loss function. The above process can be repeated for the same or different inputs until a minimum of the loss function is achieved, minimizing the training error. Several training methods are known in the art, including gradient descent, stochastic gradient descent, batch gradient descent, gradient boosting, and metaheuristic methods.

[0019] In some embodiments, each of the plurality of ML models 106 may be based on a hybrid architecture of multiple deep neural networks (DNNs). Examples of each of the plurality of ML models 106 may include, but are not limited to, a neural network model or a model based on one or more of regression method(s), instance-based method(s), regularization method(s), decision tree method(s), Bayesian method(s), clustering method(s), association rule learning, and dimensionality reduction method(s). Examples of neural network models may include, but are not limited to, an artificial neural network (ANN), a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a CNN-recurrent neural network (CNN-RNN), R-CNN, Fast R-CNN, Faster R-CNN, a residual neural network (Res-Net), a feature pyramid network (FPN), and / or combinations thereof.

[0020] The audio capturing device 108 may include suitable logic, circuitry, and / or interfaces that may be configured to capture verbal interactions (in the form of audio signals) between each of a plurality of attendees 116 and an educator of a corresponding meeting session. Examples of the audio capturing device 108 may include, but are not limited to, a recorder, an electret microphone, a dynamic microphone, a carbon microphone, a piezoelectric microphone, a fiber microphone, a (microelectromechanical electromechanical system) MEMS microphone, or other microphones known in the art. The audio capturing device 108 may be positioned proximate to the attendees and the educator to capture the verbal interactions during different meeting sessions.

[0021] Server 110 may include suitable logic, circuitry, interfaces, and code that may be configured to store received images 114 of attendees 116 associated with multiple meeting sessions. In some embodiments, server 110 may be configured to train and store each of multiple ML models 106 for each of multiple attendees 116. In some embodiments, server 110 may be configured to store content presented (or to be presented) during meeting sessions, as well as store profile information for different attendees and educators. In some embodiments, server 110 may be implemented as a cloud server capable of performing operations via web applications, cloud applications, HTTP requests, repository operations, file transfers, etc. Other examples of server 110 may include, but are not limited to, a database server, a file server, a web server, a media server, an application server, a mainframe server, a cloud server, or other types of servers. In one or more embodiments, server 110 may be implemented as multiple distributed cloud-based resources using techniques known to those skilled in the art. Those skilled in the art will appreciate that the scope of the present disclosure may not be limited to the implementation of server 110 and system 102 as separate entities. In some embodiments, the functionality of server 110 may be incorporated, in whole or at least in part, into system 102 without departing from the scope of the present disclosure.

[0022] The communications network 112 may include a communications medium that enables the system 102, the plurality of image capture devices 104, the audio capture device 108, and the server 110 to communicate with one another. The communications network 112 may be a wired or wireless communications network. Examples of the communications network 112 may include, but are not limited to, the Internet, a cloud network, a wireless fidelity (Wi-Fi) network, a personal area network (PAN), a local area network (LAN), or a metropolitan area network (MAN). The various devices in the network environment 100 may be configured to connect to the communications network 112 according to various wired and wireless communications protocols. Examples of such wired and wireless communication protocols include, but are not limited to, at least one of Transmission Control Protocol and Internet Protocol (TCP / IP), User Datagram Protocol (UDP), Hypertext Transfer Protocol (HTTP), File Transfer Protocol (FTP), ZigBee, EDGE, IEEE802.11, Light Fidelity (Li-Fi), 802.16, IEEE802.11s, IEEE802.11g, multi-hop communication, wireless access point (AP), device-to-device communication, cellular communication protocols, and Bluetooth (BT) communication protocols.

[0023] In operation, the system 102 can be configured to receive multiple images 114 of multiple attendees 116 associated with multiple meeting sessions (e.g., meeting sessions attended by different attendees in the past). The multiple images 114 can be received from multiple image capture devices 104 located at the different meeting sessions. The meeting categories of one or more of the multiple meeting sessions can be different and can correspond to at least one of the type of meeting session, the number of attendees in the meeting session, the duration of the meeting session, the average age of attendees in the meeting session, the topic of the meeting session, the experience of an educator in the meeting session, or the content presented in the meeting session. The system 102 can further detect one or more activities that each of the multiple attendees 116 can engage in during the corresponding meeting session. The one or more activities can be detected based on the received multiple images 114. In some embodiments, the one or more detected activities performed by each of the plurality of attendees 116 may relate to at least one of an action performed by the attendee, a gesture performed by the attendee, a head posture of the attendee, a body position of the attendee, lip movement of the attendee, gaze of the attendee, or a facial expression of the attendee. In other words, the one or more activities may indicate at least one of a level of concentration of the attendee or a level of interaction between the attendee and the educator. More details regarding the detection of the one or more activities are shown, for example, in FIG. 3.

[0024] In one embodiment, the system 102 can be configured to control the audio capture devices 108 to capture interactions between multiple attendees 116 and the educator of a corresponding meeting session. In some embodiments, the system 102 can receive multiple audio signals (i.e., captured by corresponding audio capture devices during different meeting sessions) directly from the server 110. Based on the captured interactions, the system 102 can further determine one or more keywords in the captured interactions. Details regarding determining one or more keywords are shown, for example, in FIG. 4.

[0025] The system 102 may be further configured to calculate an attention score for each of the plurality of attendees over a corresponding time period based on one or more detected activities associated with the corresponding attendee that occur during different meeting sessions. In another embodiment, the system 102 may be further configured to calculate an attention score for each of the plurality of attendees over a corresponding time period based on one or more detected activities associated with the corresponding attendee and the corresponding attendee's captured interactions with the educator. The attention score may indicate each attendee's attention level in the corresponding meeting session. Details of the attention level calculation are shown, for example, in FIG. 3. The system 102 may be further configured to train each of the plurality of ML models 106 for each of the plurality of attendees 116 based on the calculated attention score for each of the plurality of attendees 116 and the meeting category of the corresponding meeting session.

[0026] The system 102 can be configured to store the trained ML models 106 in a memory (i.e., memory 204 in FIG. 2 ) based on training the multiple ML models 106 and apply the multiple ML models 106 in a real-world scenario. In the real-world scenario, the system 102 can be configured to receive a first set of images of a first attendee among the multiple attendees 116. The first attendee can be associated with a first meeting session (i.e., a recent meeting session) that can be different from the multiple meeting sessions (i.e., past meeting sessions on which the multiple ML models 106 are trained). The system 102 can further detect a first set of activities that the first attendee can perform during the first meeting session over a first period of time (e.g., several minutes during the meeting session). The system 102 can further calculate a first attention score associated with the first attendee over the first period of time based on the detected first set of activities. The first attention score can indicate an attention level of the first attendee in the first meeting session. The system 102 may further apply a first machine learning (ML) model of the plurality of ML models 106 based on the calculated first attention score. The first machine learning (ML) model may be personalized for the first attendee or may be trained based on past data related to the first attendee (i.e., attention score, meeting category, recommendations, other factors such as the attendee's facial expression (or emotion), the educator's experience, the attendee's preferences / interests, the environmental conditions of the meeting session, and the attendee's response time, etc.). The system 102 may further determine a first recommendation set based on application of the first ML model to the first attendee's calculated first attention score, and may further output the determined first recommendation set for either the first attendee, the educator of the first meeting session, or the content presented during the first meeting session. Details regarding the application of the trained first ML model and recommendation set are shown, for example, in FIG. 4.

[0027] FIG. 2 is a block diagram illustrating an exemplary system for attention-determination-based meeting session control, according to an embodiment of the present disclosure. The description of FIG. 2 is provided with reference to elements in FIG. 1 . FIG. 2 illustrates a block diagram 200 of a system 102. The system 102 may include a circuit 202 that can perform operations for training a plurality of ML models 106 and further apply the plurality of trained ML models 106 for attention-determination-based meeting session control. The system 102 may further include a memory 204, an input / output (I / O) device 206, a network interface 208, one or more NN models 210, and an inference accelerator 212. The memory 204 may include the plurality of ML models 106 and one or more neural network (NN) models 210. The circuit 202 may be communicatively coupled to the memory 204, the I / O device 206, the network interface 208, and the inference accelerator 212.

[0028] The circuitry 202 may include suitable logic, circuits, and interfaces that can be configured to execute program instructions associated with different operations performed by the system 102. For example, some of the operations may include receiving a plurality of images 114, detecting one or more activities, calculating an attention score, training a plurality of ML models 106, applying the trained ML models, and determining recommendations. The circuitry 202 may include one or more specialized processing units that may be implemented as independent processors. In some embodiments, the one or more specialized processing units may be implemented as an integrated processor or a group of processors that collectively perform the functions of the one or more specialized processing units. The circuitry 202 may be implemented based on multiple processor technologies known in the art. Example implementations of the circuitry 202 may be an X86-based processor, a graphics processing unit (GPU), a reduced instruction set computing (RISC) processor, an application-specific integrated circuit (ASIC) processor, a complex instruction set computing (CISC) processor, a microcontroller, a central processing unit (CPU), and / or other control circuitry.

[0029] The memory 204 may include suitable logic, circuitry, interfaces, and / or code that may be configured to store the received images 114, the trained ML models 106, and the one or more NN models 210. The memory 204 may be further configured to store a first three-dimensional (3D) map of the meeting session, attendee focus scores, attendee interaction scores, facial expressions of each of the plurality of attendees 116, educator experience information for the meeting session, attendee profile information, meeting session environment information, the first set of images, and dashboard information. Example implementations of the memory 204 may include, but are not limited to, random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), hard disk drive (HDD), solid-state drive (SSD), CPU cache, and / or secure digital (SD) card.

[0030] The I / O devices 206 may include suitable logic, circuitry, and interfaces that may be configured to receive user input and provide output based on the received user input. User input may include, but is not limited to, a request to provide recommendations for a particular meeting session, a request to provide dashboard information, experience information about a particular educator, or profile information about a particular attendee. The I / O devices 206 may include a variety of input and output devices that may be configured to communicate with the circuitry 202. Examples of the I / O devices 206 may include, but are not limited to, a display device 206A, an audio rendering device, a touch screen, a keyboard, a mouse, a joystick, and a microphone.

[0031] The display device 206A may include suitable logic, circuitry, and interfaces that may be configured to display dashboard information, including, but not limited to, statistics regarding a particular attendee's attention score (or regarding the meeting session), or recommendations for attendees, educators, content creators, or institutions related to the meeting session. The display device 206A may be a touch screen that allows a user to provide user input via the display device 206A. The touch screen may be at least one of a resistive touch screen, a capacitive touch screen, or a thermal touch screen. The display device 206A may be implemented through multiple known technologies, such as, but not limited to, at least one of a liquid crystal display (LCD) display, a light emitting diode (LED) display, a plasma display, or an organic LED (OLED) display technology, or other display devices. According to an embodiment, the display device 206A may represent a display screen of a head-mounted device (HMD), a smart glasses device, a see-through display, a projection display, an electrochromic display, or a transparent display.

[0032] The network interface 208 may include suitable logic, circuits, and interfaces configurable to facilitate communication between the circuit 202, the plurality of image capture devices 104, the audio capture devices 108, and the server 110 over the communications network 112. The network interface 208 may be implemented using various known technologies to support wired or wireless communications between the system 102 and the communications network 112. The network interface 208 may include, but is not limited to, an antenna, a radio frequency (RF) transceiver, one or more amplifiers, a tuner, one or more oscillators, a digital signal processor, a coder-decoder (CODEC) chipset, a subscriber identity module (SIM) card, or local buffer circuitry. The network interface 208 may be configured to communicate via wireless communications with a network, such as the Internet, an intranet, or a wireless network, such as a cellular telephone network, a wireless local area network (LAN), or a metropolitan area network (MAN). The wireless communication may be configured to use one or more of a number of communication standards, protocols, and technologies, such as Global System for Mobile Communications (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (W-CDMA), Long Term Evolution (LTE), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Bluetooth, Wireless Fidelity (WiFi) (such as IEEE 802.11a, IEEE 802.11b, IEEE 802.11g, or IEEE 802.11n), Voice over Internet Protocol (VoIP), Light Fidelity (Li-Fi), Worldwide Interoperability for Microwave Access (Wi-MAX), protocols for email, instant messaging, and short message service (SMS).

[0033] Each of the one or more NN models 210 can be a computer network or a system of artificial neurons arranged in multiple layers as nodes. The multiple layers of each of the one or more NN models 210 can include an input layer, one or more hidden layers, and an output layer. Each of the multiple layers can include one or more nodes (or artificial neurons, represented by circles, for example). The output of every node in the input layer can be connected to at least one node in the hidden layer(s). Similarly, the input of each hidden layer can be connected to the output of at least one node in another layer of each of the one or more NN models 210. The output of each hidden layer can be connected to the input of at least one node in another layer of each of the one or more NN models 210. The node(s) in the final layer can receive input from at least one hidden layer and output a result. The number of layers and the number of nodes in each layer can be determined from the hyperparameters of each of the one or more NN models 210. Such hyperparameters can be set before, during, or after training of each of the one or more NN models 210 based on a training dataset (e.g., training dataset 116). Each node of the one or more NN models 210 can correspond to a mathematical function (e.g., a sigmoid function or a rectified linear unit) having a parameter set that can be adjusted during training of the network. The parameter set can include, for example, weight parameters and regularization parameters. Each node can calculate an output using a mathematical function based on one or more inputs from nodes in other layer(s) (e.g., previous layer(s)) of each of the one or more NN models 210. All or some of the nodes of each of the one or more NN models 210 can correspond to the same or different mathematical functions.

[0034] Training each of the one or more NN models 210 may include updating one or more parameters of each node based on whether the output of the final layer for a given input (from the training dataset) matches the correct result based on the loss function of each of the one or more NN models 210. The above process may be repeated for the same or different inputs until a minimum of the loss function is achieved, minimizing the training error. Several training methods are known in the art, such as gradient descent, stochastic gradient descent, batch gradient descent, gradient boosting, and metaheuristic methods.

[0035] Each of the one or more NN models 210 may include electronic data that may be implemented, for example, as a software component of an application executable on system 102. Each of the one or more NN models 210 may rely on libraries, external scripts, or other logic / instructions for execution by a device such as circuit 202. Each of the one or more NN models 210 may include code and routines configured to enable a computing device such as circuit 202 to perform one or more operations to detect one or more activities performed by multiple attendees. Additionally or alternatively, each of the one or more NN models 210 may be implemented using hardware, including a processor, a microprocessor (e.g., that performs or controls one or more operations), a field programmable gate array (FPGA), or an application-specific integrated circuit (ASIC). Alternatively, in some embodiments, each of the one or more NN models 210 may be implemented using a combination of hardware and software. Examples of the one or more NN models 210 may include, but are not limited to, a deep neural network (DNN), a convolutional neural network (CNN), an artificial neural network (ANN), a CNN+ANN, R-CNN, Fast R-CNN, Faster R-CNN, (You Only Look Once) YOLO network, a fully connected neural network, and / or combinations of such networks.

[0036] The inference accelerator 212 may include suitable logic, circuitry, interfaces, and / or code that may be configured to operate as a co-processor of the circuit 202 to accelerate computations associated with the operation of one or more NN models 210 and / or the plurality of ML models 106. For example, the inference accelerator 212 may accelerate computations of the system 102 such that one or more activities are detected in a shorter time than would normally occur without the use of the inference accelerator 212. The inference accelerator 212 may implement various acceleration techniques, such as parallelizing some or all of the operation of the one or more NN models 210 and / or the plurality of ML models 106. The inference accelerator 212 may be implemented as software, hardware, or a combination thereof. Examples of implementations of the inference accelerator 212 include, but are not limited to, a GPU, a tensor processing unit (TPU), a neuromorphic chip, a vision processing unit (VPU), a field programmable gate array (FPGA), a reduced instruction set computing (RISC) processor, an application specific integrated circuit (ASIC) processor, a complex instruction set computing (CISC) processor, a microcontroller, and / or combinations thereof.

[0037] Figure 3 illustrates an example operation for training an ML model for each of multiple attendees associated with multiple meeting sessions, according to an embodiment of the present disclosure. The description of Figure 3 is provided with reference to elements in Figures 1 and 2. Figure 3 illustrates a block diagram 300 illustrating example operations 302A-302J described herein. The example operations illustrated in block diagram 300 may begin at 302A and may be performed by any computer system, apparatus, or device, such as system 102 of Figure 1 or circuit 202 of Figure 2. While the example operations associated with one or more blocks of block diagram 300 are illustrated as discrete blocks, they may be divided into additional blocks, combined into fewer blocks, or eliminated depending on the particular implementation.

[0038] At 302A, a data capture operation may be performed. In the data capture operation, the circuitry 202 may be configured to receive multiple images 114 of multiple attendees 116 associated with multiple meeting sessions. In some embodiments, the circuitry 202 may control multiple image capture devices 104 over a period of time (e.g., days, weeks, months, or years) to capture multiple images 114 of the multiple attendees 116. The circuitry 202 may further receive multiple images 114 captured from the multiple image capture devices 104. In some embodiments, the circuitry 202 may receive multiple images 114 of the multiple attendees 116 (i.e., attendees who attended the corresponding meeting session) from the server 110.

[0039] In some embodiments, each meeting session (e.g., a lecture shown in FIG. 1 ) may include multiple image capture devices 104 positioned at different locations in the meeting session to capture images of attendees from different perspectives. In such cases, the multiple image capture devices 104 may include, but are not limited to, a first image capture device, a second image capture device, and a third image capture device. The first image capture device may capture in an axial image plane (i.e., on the x-y axis), the second image capture device may capture in an orthogonal image plane (i.e., on the y-z axis), and the third image capture device may capture in a sagittal image plane (i.e., on the x-z axis). In some embodiments, the second and third image capture devices may be optional, and the first image capture device may capture images of attendees present in a particular meeting session. In some embodiments, the first image capture device may include a depth sensor (e.g., an RGBD camera) to capture images of attendees along with depth information (indicating the distance between the first image capture device and each attendee).

[0040] In some embodiments, the meeting session may be a classroom-based (i.e., offline-based) session, such as that shown in FIG. 1 . In such cases, examples of the meeting session may include, but are not limited to, an educational class, a physical meeting of a professional organization, a physical conference, a physical seminar, or a physical training room. In other embodiments, the meeting session may be an online-based session, such as, but not limited to, an online class, an online meeting session, an online training session, a web conference, or a webinar. In such cases, multiple image capture devices 104 (e.g., CCTVs) may be located within a room (or any physical enclosure) from which attendees can attend the corresponding online meeting session. In some cases, multiple image capture devices 104 (e.g., webcams) may be integrated into (or built into or coupled to) a computing device (e.g., a mobile phone, personal computer, laptop, etc.) from which attendees can attend the corresponding online meeting session.

[0041] In some embodiments, one or more of the plurality of meeting sessions may have different meeting categories. The meeting category may correspond to at least one of the following: a type of meeting session, a number of attendees in the meeting session, a duration of the meeting session, an average age of attendees in the meeting session, a topic of the meeting session, an experience of an educator in the meeting session, or content presented in the meeting session. In some embodiments, examples of different types of meeting sessions may include, but are not limited to, a classroom-based session with an educator, a classroom-based session without an educator, an online-based session with an educator, and an online-based session without an educator. The number of attendees may indicate the number of attendees present in a particular meeting session, such as a low-intensity meeting (i.e., 1-10 attendees or less than 30% meeting intensity), a medium-intensity meeting (i.e., 11-20 attendees or 31-60% meeting intensity), or a high-intensity meeting (i.e., 20 or more attendees or greater than 60% meeting intensity). The duration of a meeting session may indicate whether the meeting session is short (e.g., less than 30 minutes) or long (e.g., more than 1 hour). The average age of attendees may indicate the age group or educational year of attendees (e.g., preschool attendees, elementary standard attendees, high school standard attendees, first-year graduating attendees, final-year graduating attendees, 3-6 year old attendees, 12-17 year old attendees, 21-30 year old attendees, or 35+ year old attendees). The topic of a meeting session may indicate the agenda of the meeting session (e.g., but not limited to, the subject of the meeting session, a particular chapter of a curriculum, a technology or industry topic, a particular plan to be discussed during the meeting session, or a particular issue to be discussed during the meeting session). The experience of an educator may indicate the number of years of teaching (or training) experience of an educator (or teacher, trainer, or instructor), such as less than 1 year of experience, 2-5 years of experience, 6-10 years of experience, or more than 10 years of experience.The content presented in the meeting session can indicate whether the content presented relates to theoretical content, includes multiple practical examples, or is interactive content.

[0042] In one embodiment, the disclosed system 102 can consider images of attendees in different categories of meeting sessions because different meeting categories may influence or affect the attention level of the attendees. Accordingly, the system 102 can track the attention scores of particular attendees in different categories of meeting sessions over a particular period of time (e.g., months or years) to effectively train the personalized ML model using diverse and robust training data for the particular attendees.

[0043] At 302B, a 3D map generation operation can be performed. In the 3D map generation operation, the circuit 202 can be configured to generate a first three-dimensional (3D) map of a corresponding meeting session that includes at least one of the plurality of attendees 116. The generation of the first 3D map can correspond to three-dimensionally profiling the corresponding meeting session (i.e., including at least one of the plurality of attendees 116) to map objects or attendees that exist in the real world. The circuit 202 can generate the first 3D map of the meeting session based on a plurality of captured images 114 of the meeting session that includes the corresponding attendees. In one embodiment, the system 102 can include a video (or imaging) multiplexer 304 that can combine the plurality of images 114 of a particular meeting session captured by the first image capture sensor, the second image capture sensor, and the third image capture sensor. The system 102 can further control the multiplexer 304 to generate the first 3D map of the meeting session. In an embodiment, the system 102 can utilize the depth information (eg, captured by a depth sensor in an image capture device) to generate a first 3D map of the meeting session including the corresponding attendees.

[0044] At 302C, an activity detection operation may be performed. In the activity detection operation, the circuit 202 may be configured to detect one or more activities that each of the plurality of attendees 116 may perform during the corresponding meeting session. The circuit 202 may detect one or more activities of a particular attendee based on the plurality of received images 114 related to the corresponding meeting session in which the attendee is present. In one embodiment, the circuit 202 may be configured to apply one or more neural network (NN) models 210 to the plurality of received images 114 or the generated first 3D map of the meeting session to detect one or more activities of the corresponding attendee over a period of time (e.g., over minutes or hours during the meeting session).

[0045] In some embodiments, one or more activities performed by each of the plurality of attendees 116 may be associated with at least one of an action performed by the attendee (as shown in FIG. 3 ), a gesture performed by the attendee, a head posture of the attendee, a body position of the attendee, lip movement of the attendee, gaze of the attendee, or facial expression of the attendee. Examples of the one or more activities associated with an action performed by an attendee may include, but are not limited to, writing, typing, talking, asking a question, yawning, carrying an object, and performing a group activity. One or more activities associated with a gesture performed by an attendee may indicate whether the attendee is making a predetermined gesture to perform an action, such as (but not limited to) raising their hand to request permission to go outside for a break during a meeting session or pointing at something / someone during a particular period. One or more activities associated with lip movement of the attendee may indicate whether the attendee is reading aloud or talking to another attendee. One or more activities associated with the body position of the attendee may indicate whether the attendee is sitting or standing during the period. The one or more activities related to the attendee's gaze can indicate whether the attendee is focused during the meeting session. The one or more activities related to the attendee's head posture can indicate whether the attendee is looking left, looking right, present, absent, or with their head down during the meeting session.

[0046] In some embodiments, the circuit 202 can be configured to detect one or more attendee activities across a set of time slots within the time period. The circuit 202 can detect at least one of the one or more activities (i.e., an activity performed by the multiple attendees 116) in each of the set of time slots within the time period. For example, if the time period is 10 minutes, each of the set of time slots can be one minute or a few seconds. In some embodiments, each of the set of time slots can correspond to a specific number of image frames in which specific attendee activities are detected using the corresponding multiple images 114. The circuit 202 can generate a timeline including the set of time slots for a specific meeting session and the detected activity of each of the multiple attendees 116 in each time slot, as shown in Table 1 below. TIFF2025510492000002.tif35155 Table 1: An exemplary timeline of the detected activities of different attendees of the corresponding meeting session. The parameters in Table 1 are as follows: AQ shows attendees answering questions in a meeting session, F indicates that the attendee is focused on the content of the meeting session, RH indicates that an attendee has raised their hand in a meeting session, R indicates that an attendee is reading in a meeting session, W indicates that the attendee is writing or typing in the meeting session, T indicates that the attendee is talking to someone in a meeting session, E indicates that the attendee is not present in their assigned seat in the meeting session (i.e., out of place); S indicates that an attendee is standing in a meeting session, FM corresponds to frame not detected or activity not detected.

[0047] In some embodiments, the circuit 202 can determine the location of each attendee within a meeting session (e.g., in a physical classroom-based meeting session) based on the generated first 3D map of the meeting session. The circuit 202 can associate or tag a specific seat with the attendee's location based on the generated first 3D map of the meeting session. Tagging an attendee with a specific seat can help recognize the attendee during the remainder of the meeting session and identify activities performed by the same attendee while sitting in the same seat tagged to the attendee. Similarly, the circuit 202 can determine the location of each of the multiple attendees 116 in the corresponding meeting session based on the received multiple images 114 and the determined 3D map of each meeting session. Furthermore, the circuit 202 can associate each of the multiple attendees 116 with a different seat in the respective meeting session. Based on such seat tagging, the circuit 202 can determine in which location (front, back, center, or corner) of the meeting session a particular attendee sat and performed different activities while attending the respective meeting session. Thus, different actions (or activities or interactions) can be associated with corresponding attendees of different meeting sessions based on the seat tagging performed by the disclosed system 102 .

[0048] At 302D, a concentration score calculation operation may be performed. In the concentration score calculation operation, the circuit 202 may be configured to calculate a concentration score for each of the plurality of attendees 116 for a corresponding meeting session in which each attendee is present. The circuit 202 may calculate the concentration score for each of the plurality of attendees 116 based on one or more detected activities associated with the corresponding attendee. In other words, the circuit 202 may calculate a concentration score for a particular attendee for a particular meeting session. Specifically, the concentration score may be calculated based on the detection of a first set of one or more activities. Such a first set of activities may include, but is not limited to, actions performed by the attendee (e.g., writing, typing, talking, yawning, etc.), the attendee's gaze, the attendee's head posture, the attendee's body position, and the attendee's lip movement (e.g., reading aloud or talking to other attendees).

[0049] In one embodiment, the circuit 202 can be configured to determine a first duration of at least one activity of the one or more activities performed by each of the multiple attendees during the corresponding meeting session. Specifically, the circuit 202 can be configured to determine a first duration of each of a first set of activities. The circuit 202 can be configured to assign a score to each of the first set of activities based on the determined first duration. In one embodiment, the circuit 202 can be configured to assign scores or points (such as, but not limited to, writing-related points, reading-related points, gaze-related points, or speech-related points) for different activities to each of the multiple attendees 116 based on the plurality of captured images 114 of the corresponding meeting session. In one embodiment, the determined first duration can indicate the amount of time the attendee spent on the different activities, which can be further used to tag the attendee as an “attention seeker” and / or an “attentive attendee.”

[0050] The circuit 202 can be configured to determine the intensity (i.e., number of attendees) of each meeting session of the plurality of meeting sessions for assigning reading-related points. The circuit 202 can further compare the determined meeting session intensity with a threshold intensity (e.g., 30% intensity). If the determined intensity is higher than the threshold intensity, the system 102 can apply a read formula to assign reading-related points to each of the plurality of attendees 116 that may be present in the meeting session. The read-related formula can be applied by the following equation (1): TIFF2025510492000003.tif9150(1) where: R corresponds to the reading-related points or score, C corresponds to the type of meeting session, k corresponds to the duration of the reading activity in seconds / minutes (i.e., the first duration); S corresponds to the intensity of the meeting session, C1 supports webinar meeting sessions with educators, C2 supports webinar meeting sessions without an educator, C3 corresponds to a class meeting session with an educator, C4 supports class meeting sessions without an educator, C5 corresponds to an interaction / activity meeting session with an educator, C6 corresponds to an interaction / activity meeting session without an educator, Cn=0 or 1. When C1=1, C2=C3=C4=C5=C6=0, P1, P2, P3, P4, P5 and P6 correspond to weight variables having values ​​between 0 and 1.

[0051] In one embodiment, if the percentage (i.e., intensity) of attendees (A) present / available in the meeting session is greater than 50% and the duration of an attendee's reading activity is greater than or equal to 10 seconds and less than or equal to 40 seconds, each of the plurality of attendees 116 reading aloud in a particular meeting session may be assigned a reading-related point of "0.9," while the remaining attendees (i.e., other than the attendee who was reading aloud) may be assigned a reading-related point of "0.1." In another embodiment, if the percentage (i.e., intensity) of attendees (A) present / available in the meeting session is greater than 50% and the duration of an attendee's reading activity is greater than 40 seconds, each of the plurality of attendees 116 reading aloud may be assigned a reading-related point of "1," while the remaining attendees (i.e., other than the attendee who was reading aloud) may be assigned a reading-related point of "0." In another embodiment, if the percentage of attendees (A) present / able to attend the meeting session is greater than 50% and the duration of an attendee's reading activity is less than 10 seconds, each of the attendees 116 who are reading aloud may be assigned 0 reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may also be assigned 0 reading-related points.

[0052] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' reading activity is greater than 40 seconds, each of the plurality of attendees 116 reading aloud may be assigned "0.5" reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may also be assigned "0.5" reading-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' reading activity is less than 10 seconds, each of the plurality of attendees 116 reading aloud may be assigned "0" reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may also be assigned "0" reading points. In one embodiment, if the percentage of attendees (A) present / able to attend the meeting session is less than 25% and the duration of an attendee's reading activity is 10 seconds or more but less than 40 seconds, each of the attendees 116 who are reading aloud may be assigned a reading-related point of "0", while the remaining attendees (i.e., other than the attendee who was reading aloud) may also be assigned a reading-related point of 0.

[0053] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of the attendees' reading activity is greater than 10 seconds and less than 40 seconds, each of the plurality of attendees 116 reading aloud may be assigned 0.8 reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may be assigned 0.2 reading-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of the attendees' reading activity is greater than 40 seconds, each of the plurality of attendees 116 reading aloud may be assigned 0.6 reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may be assigned 0.4 reading-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% but less than 50%, and the duration of the attendees' reading activity is less than 10 seconds, each of the attendees 116 who are reading aloud may be assigned 0 reading-related points, while the remaining attendees (i.e., other than the attendee who was reading aloud) may also be assigned 0 reading-related points. Table 2 below illustrates an exemplary relationship between the intensity of the meeting session (i.e., the number of attendees (A)), the duration of the reading activity, and the reading-related points. TIFF2025510492000004.tif67155 Table 2: Exemplary relationships between meeting session intensity, duration of recitation activities and recitation related points.

[0054] In one embodiment, to assign writing-related points, the circuit 202 can be configured to determine an intensity of each meeting session of the plurality of meeting sessions. The circuit 202 can further compare the determined intensity of the meeting session with a threshold intensity (e.g., 30% intensity). If the determined intensity is higher than the threshold intensity, the system 102 can apply a write formula to assign writing-related points to each of the plurality of attendees 116 that may be present in the meeting session. The write formula can be applied according to the following equation (2): TIFF2025510492000005.tif11157(2) where: W corresponds to writing-related points or scores, C corresponds to the type of meeting session, k corresponds to the duration of the writing activity in seconds / minutes (i.e., the first duration); S corresponds to the intensity of the meeting session, C1 supports webinar meeting sessions with educators, C2 supports webinar meeting sessions without an educator, C3 corresponds to a class meeting session with an educator, C4 supports class meeting sessions without an educator, C5 corresponds to an interaction / activity meeting session with an educator, C6 corresponds to an interaction / activity meeting session without an educator, Cn=0 or 1. When C1=1, C2=C3=C4=C5=C6=0, P1, P2, P3, P4, P5 and P6 correspond to weight variables having values ​​between 0 and 1.

[0055] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 50% and the duration of an attendee's writing activity is greater than or equal to 3 seconds but less than 15 seconds, each of the attendees 116 taking notes may be assigned a writing-related point of "0.9," while the remaining attendees (i.e., other than the attendee who was taking notes) may be assigned a writing-related point of "0.1." In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 50% and the duration of an attendee's writing activity is greater than 15 seconds, each of the attendees 116 taking notes may be assigned a writing-related point of "1.0," while the remaining attendees (i.e., other than the attendee who was taking notes) may be assigned a writing-related point of 0. In another embodiment, if the percentage of attendees (A) who are present / able to attend the meeting session is greater than 50% and the duration of an attendee's writing activity is less than 3 seconds, each of the attendees 116 who are taking notes may be assigned 0 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may also be assigned 0 writing-related points.

[0056] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' writing activity is greater than 15 seconds, each of the attendees 116 taking notes may be assigned 0.5 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may also be assigned 0.5 writing-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' writing activity is less than 3 seconds, each of the attendees 116 taking notes may be assigned 0 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may also be assigned 0 writing-related points. In one embodiment, if the percentage of attendees (A) present / able to attend the meeting session is less than 25% and the duration of an attendee's writing activity is greater than or equal to 3 seconds and less than or equal to 15 seconds, each of the attendees 116 who are taking notes may be assigned 0 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may also be assigned 0 writing-related points.

[0057] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of an attendee's writing activity is greater than 3 seconds and less than 15 seconds, each of the attendees 116 taking notes may be assigned 0.8 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may be assigned 0.2 writing-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of an attendee's writing activity is greater than 15 seconds, each of the attendees 116 taking notes may be assigned 0.6 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may be assigned 0.4 writing-related points. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% but less than 50%, and the duration of an attendee's writing activity is less than 3 seconds, each of the attendees 116 who are taking notes may be assigned 0 writing-related points, while the remaining attendees (i.e., other than the attendee who was taking notes) may also be assigned 0 writing-related points. Table 3 below illustrates an exemplary relationship between the intensity of the meeting session (i.e., the number of attendees (A)), the duration of the writing activity, and the writing-related points. TIFF2025510492000006.tif67155 Table 3: Exemplary relationships between meeting session intensity, duration of writing activities and writing-related points.

[0058] In one embodiment, to assign gaze-related points, the circuit 202 can be configured to determine an intensity of each meeting session of the plurality of meeting sessions. The circuit 202 can further compare the determined intensity of the meeting session with a threshold intensity (e.g., 30% intensity). If the determined intensity is higher than the threshold intensity, the system 102 can apply a gaze formula to assign gaze-related points to each of the plurality of attendees 116 that may be present in the meeting session. The gaze formula can be applied according to the following equation (3): TIFF2025510492000007.tif11157(3) where: X corresponds to a gaze-related point or score, C corresponds to the type of meeting session, k corresponds to the duration of the gaze activity in seconds / minutes (i.e., the first duration), S corresponds to the intensity of the meeting session, C1 supports webinar meeting sessions with educators, C2 supports webinar meeting sessions without an educator, C3 corresponds to a class meeting session with an educator, C4 supports class meeting sessions without an educator, C5 corresponds to an interaction / activity meeting session with an educator, C6 corresponds to an interaction / activity meeting session without an educator, Cn=0 or 1. When C1=1, C2=C3=C4=C5=C6=0, P1, P2, P3, P4, P5 and P6 correspond to weight variables having values ​​between 0 and 1.

[0059] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 50% and the duration of the attendees' gaze activity is greater than or equal to 30 seconds but less than 90 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of "0.9", while the remaining attendees (i.e., other than the attendee who was gazing) may be assigned a gaze-related point of "0.1". In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 50% and the duration of the attendees' gaze activity is greater than 90 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of "1.0", while the remaining attendees (i.e., other than the attendee who was gazing) may be assigned a gaze-related point of 0. In another embodiment, if the percentage of attendees (A) who are present / able to attend the meeting session is greater than 50% and the duration of the attendees' gaze activity is less than 30 seconds, each of the attendees 116 who are gazing can be assigned a gaze-related point of 0, while the remaining attendees (i.e., other than the attendee who was gazing) can also be assigned a gaze-related point of 0.

[0060] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' gaze activity is greater than 90 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of "0.5", while the remaining attendees (i.e., attendees other than the attendee who were gazing) may also be assigned a gaze-related point of "0.5". In another embodiment, if the percentage of attendees (A) present / available in the meeting session is less than 25% and the duration of the attendees' gaze activity is less than 30 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of 0, while the remaining attendees (i.e., attendees other than the attendee who were gazing) may also be assigned a gaze-related point of 0. In one embodiment, if the percentage of attendees (A) present / able to attend the meeting session is less than 25% and the duration of the attendees' gaze activity is greater than or equal to 30 seconds but less than 90 seconds, each of the attendees 116 who are gazing can be assigned a gaze-related point of 0, while the remaining attendees (i.e., other than the attendee who was gazing) can also be assigned a gaze-related point of 0.

[0061] In one embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of the attendees' gaze activity is greater than 30 seconds and less than 90 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of 0, while the remaining attendees (i.e., attendees other than the attendee who were gazing) may also be assigned a gaze-related point of 0. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% and less than 50%, and the duration of the attendees' gaze activity is greater than 90 seconds, each of the plurality of attendees 116 who are gazing may be assigned a gaze-related point of 0.5, while the remaining attendees (i.e., attendees other than the attendee who were gazing) may also be assigned a gaze-related point of 0.5. In another embodiment, if the percentage of attendees (A) present / available in the meeting session is greater than 25% but less than 50%, and the duration of the attendees' gaze activity is less than 30 seconds, each of the attendees 116 who are gazing may be assigned a gaze-related point of 0, while the remaining attendees (i.e., other than the attendee who was gazing) may also be assigned a gaze-related point of 0. Table 4 below illustrates an exemplary relationship between the intensity of the meeting session (i.e., the number of attendees (A)), the duration of the gaze activity, and the gaze-related points. TIFF2025510492000008.tif67155 Table 4: Exemplary relationships between meeting session intensity, duration of gaze activity and gaze-related points.

[0062] In some embodiments, the circuit 202 can be configured to calculate a concentration score based on the reading-related points, the writing-related points, and the gaze-related points. Specifically, the concentration score for each of the multiple attendees 116 can be calculated based on the sum of the reading-related points, the writing-related points, and the gaze-related points assigned to the corresponding attendee of the meeting session. The circuit 202 can further store the calculated concentration score for a particular attendee of a particular meeting session in the memory 204.

[0063] At 302E, an interaction score calculation operation may be performed. In the interaction score calculation operation, the circuit 202 may be configured to calculate an interaction score for each of the plurality of attendees 116 for activities performed in a corresponding meeting session. The circuit 202 may calculate an interaction score for each of the plurality of attendees 116 based on one or more detected activities performed by the corresponding attendee in different meeting sessions. In other words, the circuit 202 may calculate an interaction score for a particular attendee in a particular meeting session. Specifically, the interaction score may be calculated based on the detection of a second set of one or more activities. Such a second set of activities may include, but is not limited to, gestures performed by the attendee (such as raising a hand or joining an activity), gestures performed by the attendee (such as sitting or standing), and lip movements of the attendee (such as when answering a question).

[0064] As an example, if the percentage of attendees (A) present / available in a meeting session is higher than 30% and an attendee raises their hand to speak, the attendee who raised their hand to speak may be assigned 0.5 interaction points, while the remaining attendees may be assigned 0.1 interaction points. If an attendee only speaks and does not raise their hand or stand, the corresponding attendee may be assigned 0.05 interaction points, while the remaining attendees may be assigned 0.2 interaction points. Each attendee's interaction score (i.e., a score referred to as "RA") may be further based on the assigned interaction points. Table 5 below shows an exemplary relationship between the intensity of a meeting session (i.e., the number of attendees (A)), the activities performed by the attendees, and the interaction points. TIFF2025510492000009.tif56155 Table 5: Exemplary relationships between meeting session intensity, activities performed by attendees and interaction points.

[0065] The circuit 202 may be further configured to calculate a median of the plurality of interaction scores / points calculated for all attendees of a particular meeting session (e.g., a particular class). M ) to each attendee's interaction score / points to calculate each attendee's final interaction score. This calculation can be performed to normalize the statistical attention curve (i.e., the curve formed by the calculated interaction scores for all attendees) to reduce any bias in the meeting session due to the absence of attendees who are negatively affected due to a lack of high-performing attendees / individuals in the meeting session. In one embodiment, the calculation of the final interaction score (i.e., the sum of a particular attendee's interaction score and the calculated median) can be based on the attendee's intensity in the particular meeting session. For example, if the current intensity of the meeting session is less than 30% of the total intensity, the circuit 202 can not consider the calculated median for calculating the final interaction score. In such a case, the final interaction score can be the same as each attendee's calculated interaction score. In another example, if the current intensity of the meeting session is equal to or greater than 30% of the total intensity, the circuit 202 can calculate the median (RA) based on Table 6 below. M value). TIFF2025510492000010.tif34153 Table 6: Position of interaction scores relative to the median and RA M An exemplary relationship between values.

[0066] At 302F, an information determining operation may be performed. In the information determining operation, the circuit 202 may be configured to determine information associated with the plurality of attendees 116, information associated with the plurality of educators of the plurality of meeting sessions, and information associated with the plurality of meeting sessions. In some embodiments, the circuit 202 may determine experience information that may be associated with each of the plurality of educators of a corresponding meeting session of the plurality of meeting sessions. The experience information may indicate at least one of experience, evaluation, achievements, or feedback associated with each educator of the corresponding meeting session. The circuit 202 may further be configured to store the determined experience information in the memory 204. In some embodiments, the circuit 202 may retrieve experience information about a particular educator from the server 110 or a memory to determine information about the educator. The experience information about the educator may be indicative of the educator's ability to effectively present a particular topic or content to best engage attendees.

[0067] The circuit 202 can be further configured to determine content information related to content to be presented during each of the plurality of meeting sessions. The determined content information can indicate, but is not limited to, the type of content (theoretical, interactive, content with practical / logical / real-time examples and challenges, whether it is fair to new learners), the duration of the content, the subject matter related to the content, the complexity of the content, the interactivity of the content, etc. The circuit 202 can store the determined content information in the memory 204. In some embodiments, the circuit 202 can retrieve content information related to particular content from the server 110 or memory to determine the content information related to the content to be presented during a particular meeting session.

[0068] The circuit 202 can be further configured to retrieve profile information associated with each of the plurality of attendees 116. The profile information can indicate preferences or interests in topics or content related to the corresponding meeting session. In some embodiments, the profile information can further indicate health information associated with the corresponding attendee during a particular meeting session. For example, if the attendee is a student and prefers subjects such as math and science but not social studies, the profile information can indicate that the student may be interested in math and science but not social studies. Thus, the student's attention level while attending math and science classes may be higher than the student's attention level while attending social studies classes. Similarly, the system 102 can determine, based on the health information, whether the attendee suffered from any particular health issues (such as a cold, headache, or fever) while attending the meeting session that resulted in poor attention during the meeting session. In some embodiments, the circuit 202 can retrieve profile information for a particular attendee from the server 110 or a memory to determine the profile information.

[0069] The circuit 202 may be further configured to determine environmental information associated with the geolocation of the educator of each meeting session or at least one of the multiple attendees 116 of each meeting session. The environmental information may indicate weather information associated with the geolocation of the educator of each meeting session or at least one of the multiple attendees 116 of each meeting session. Weather at a geolocation (e.g., heavy rain / wind / storm, or excessive heat / cold) may also affect the attention level of the multiple attendees 116. The environmental information may further indicate indoor / outdoor lighting conditions, the audibility of the teacher, lighting parameters associated with the device on which content is being presented, etc. Similarly, poor lighting conditions or acoustic issues during a meeting session may also affect the attention level of a particular attendee. In some embodiments, the circuit 202 may receive environmental information about the meeting session from one or more sensors (such as, but not limited to, a temperature sensor, a wind sensor, a rain sensor, a lighting sensor, or an audio sensor) located at different locations in the meeting session. The circuit 202 may be further configured to store the determined weather information in the memory 204. In some embodiments, the circuitry 202 can determine the environmental information by retrieving environmental information for a particular meeting session from the server 110 or from a memory.

[0070] At 302G, an expression determination operation may be performed. In the expression determination operation, the circuitry 202 may be configured to determine, based on the received plurality of images 114, the facial expression (or emotion) of each of the plurality of attendees 116 in the corresponding meeting session. In one embodiment, the system 102 may be configured to apply one or more NN models 210 to each of the plurality of images 114 to determine the facial expression of each of the plurality of attendees 116 in the corresponding meeting session. The system 102 may then calculate an expression parameter (E) with a value for each of the plurality of attendees 116 based on the determined facial expressions. N The circuit 202 may be further configured to initialize the facial expression parameters (E NThe circuit 202 may be further configured to determine a first facial expression that can be determined for a majority of the attendees from the plurality of attendees 116 of a particular meeting session to initialize the facial expression parameter (E) of the attendees of the meeting session with a value. For example, if the total number of the plurality of attendees 116 of the meeting session is "10" and the determined facial expressions of "7" attendees are "smile", the first facial expression may be "smile". The circuit 202 may be configured to initialize the facial expression parameter (E) of the attendees of the meeting session if the determined facial expression of the corresponding attendee is the same as the first facial expression. N ) can be configured to be initialized with a value of "1". Otherwise, the facial expression parameters (E N ) can be initialized with a value of "0". The circuit 202 calculates the facial expression parameters (E N ) may be configured to store an initialization value in the memory 204. Further examples of facial expressions may include, but are not limited to, happy, sad, impressed, surprised, angry, neutral, amazement, shock, stress, boredom, calm, excited, confused, disgust, or fear.

[0071] At 302H, an object detection operation may be performed. In the object detection operation, the circuit 202 may be configured to detect one or more objects associated with the plurality of attendees in the received plurality of images 114 for different meeting sessions in which the corresponding attendees are present. In another embodiment, the circuit 202 may be configured to apply one or more neural network (NN) models to the received plurality of images 114 to detect one or more objects associated with the plurality of attendees 116 in the received plurality of images 114. Such objects may be animated objects that may be held by or placed near the plurality of attendees 116 of the corresponding meeting session. Such objects may include, but are not limited to, pens, paper, notebooks (for writing), water bottles, toys, ornaments, electrical devices, and communication devices. The circuit 202 may be configured to store information regarding the one or more detected objects associated with each of the plurality of attendees 116 in the memory 204. The system 102 may also take into account objects held by or near attendees, as certain objects may distract certain attendees and reduce their attention levels (e.g., toys may distract young children, decorations / paintings may distract attendees, personal cell phones may distract attendees, etc.).

[0072] At 302I, an attention score calculation operation may be performed. In the attention score calculation operation, the circuit 202 may be configured to calculate an attention score for each of the plurality of attendees 116 for each meeting session (i.e., based on the activity detected for each attendee over a corresponding time period). The attention score may indicate the attention level of each attendee in the corresponding meeting session. The circuit 202 may be further configured to calculate an attention score for each of the plurality of attendees 116 based on the calculated focus score and the calculated interaction score of the corresponding attendee for each activity detected in the different meeting sessions. As an example, the calculation of the attention score based on the calculated focus score and the calculated interaction score may be performed according to the following equation (4): AS=FS * (1-Z)+IS * Z (4) where: AS corresponds to the attention score, FS corresponds to the concentration score, IS corresponds to the interaction score, Z is a multiplication factor that may depend on the type of meeting session, where 0≦Z≦1.

[0073] In one embodiment, an attendee may not be present (i.e., out of place) in their designated / tagged seat in the meeting session, as indicated by the time slot tagged as "E" in the timeline (i.e., the timeline shown in Table 1). In such a scenario, the calculation of the attention score may be performed according to the following equation (5): AS=FS * (1-Z)+IS * ZE (5) where: AS corresponds to the attention score, Supports FS and concentration scores. IS corresponds to the interaction score, Z corresponds to a multiplication factor, 0≦Z≦1; E corresponds to an out-of-place score that represents the likelihood that an attendee will not be present in their assigned seat (ie, out of place) in the corresponding meeting session.

[0074] In one embodiment, the circuit 202 can be configured to calculate an out-of-place score for a particular attendee based on the intensity of the meeting session and the duration the particular attendee is out of their designated (or otherwise identified or tagged) seat. The circuit 202 can further be configured to calculate a final attention score based on the focus score, the interaction score, and the out-of-place score as per equation (5). For example, if the current intensity of the meeting session is less than 30% of the total intensity, the circuit 202 can disregard the out-of-place score for calculating the final interaction score. In such a case, the final interaction score can be the same as the interaction score calculated for each attendee according to equation (4). In another example, if the current intensity of the meeting session is 30% or more of the total intensity, an out-of-place score (“E”) can be calculated based on the intensity of the meeting session and the duration of the out-of-place time as per Table 7 shown below. TIFF2025510492000011.tif51155 Table 7: Exemplary relationships between out-of-place score, attendance (intensity) and duration of being out-of-place.

[0075] In one embodiment, the circuitry 202 can calculate an interaction score for each attendee for each meeting session in which the attendee is present. The attendee's interaction score can be calculated based on activity detected over a defined period of time in each meeting session (i.e., the period described in 302C, for example). Thus, the disclosed system 102 can calculate multiple attention scores for each attendee for different meeting sessions in which the attendee is present, where one or more of the attendee's meeting sessions can be from different meeting categories (i.e., the meeting categories described in 302A, for example). The multiple attention scores can indicate a suitable pattern or variation of a particular candidate's attentiveness across different categories of meeting sessions (i.e., different types, different intensities, different durations, different ages, different topics / content, different educator experiences).

[0076] In one embodiment, the circuitry 202 can be configured to calculate an attention score for each of the plurality of attendees 116 based on the calculated concentration score, the calculated interaction score of the corresponding attendee, and stored experience information associated with the educator of the respective meeting session. In this case, the calculation of the attention score can be based on the calculated concentration score, the calculated interaction score, and the determined experience information and can be performed using the following equation (6): AS=(FS * (1-Z)+IS * ZE) / (R T +R L ) (6) where: AS corresponds to the attention score, FS corresponds to the concentration score, IS corresponds to the interaction score, Z is a multiplication factor, 0≦Z≦1; E corresponds to the time that an attendee may not be present in their assigned seat, R Tcorresponds to the determined empirical information, R L corresponds to a rating provided by an attendee for the content presented in the corresponding meeting session.

[0077] For example, attendees may be more likely to pay attention to educators and / or content with higher ratings, and therefore, one can attempt to equalize attention scores for less experienced educators (teachers) or educators teaching difficult subjects based on dividing low attention scores by low educator / content ratings, or high attention scores by high educator / content ratings.

[0078] In another embodiment, the circuit 202 may be configured to use the calculated concentration score, the calculated interaction score, and the facial expression parameters (E N ) can be configured to calculate an attention score based on the initialization value of . Such calculation of the attention score can be performed using the following equation (7): AS=FS * (1-Z)+IS * Z+E N (7) where: AS corresponds to the attention score, FS corresponds to the concentration score, IS corresponds to the interaction score, Z is a multiplication factor, 0≦Z≦1; E N corresponds to the facial expression parameter, and E N =0 or 1.

[0079] In another embodiment, the circuit 202 can be configured to calculate an attention score based on profile information associated with each of the plurality of attendees 116, determined environmental information (e.g., as described at 302F), a calculated concentration score, and a calculated interaction score for the corresponding attendee. In another embodiment, the circuit 202 can be configured to calculate an attention score associated with each of the plurality of attendees 116 based on the calculated concentration score, the calculated interaction score, and one or more detected objects (e.g., as described at 302H) associated with the plurality of attendees 116. Thus, in addition to considering the attendee's concentration and interaction, the disclosed system 102 can consider different influencing factors (such as, but not limited to, educator experience information, attendee profile information, content information, environmental information, and information about objects present in the meeting session) that may affect the attention score calculated for an attendee for different activities performed during a respective meeting session. In one embodiment, the disclosed system 102 can fine-tune the attention score (i.e., the attention score calculated according to equation (4) based on the concentration score and interaction score) using different influence factors on attendee attention (i.e., using equation (5), equation (6), and equation (7)).

[0080] At 302J, an ML model training operation can be performed. In the ML model training operation, the circuit 202 can be configured to train a plurality of ML models 106 for a plurality of attendees 116 based on the calculated attention score of each of the plurality of attendees 116 and the meeting category of the corresponding meeting session. Specifically, the circuit 202 can be configured to train a plurality of ML models 106 for the plurality of attendees 116. Each ML model of the plurality of ML models 106 can be personalized for each attendee. Each personalized ML model can be trained based on (or store attention scores for) the corresponding attendee's various attention scores determined from different meeting sessions (of different categories). Furthermore, for each attention score, the personalized ML model can also be trained based on corresponding influencing factors (i.e., educator experience, content, attendee profile, environmental factors, or nearby objects) identified from different meeting sessions previously attended by the corresponding attendee.

[0081] In one embodiment, a personalized ML model can be trained for or recognize different seating locations (i.e., front, back, center, corner, near a window, or near a door) where a particular attendee may be seated while attending a corresponding (at least physical class-based) meeting session. An attendee's location can also affect the corresponding attendee's attention level / score. For example, a first attendee sitting closer to the educator may be more attentive than a second attendee sitting further away from the educator. Thus, the first attendee's attention score may be higher than the second attendee's attention score. The circuit 202 can be further configured to train a plurality of ML models 106 for each of the plurality of attendees 116 based on the calculated attention score and the determined location of each of the plurality of attendees 504 in the corresponding meeting session. As described, for example, in FIG. 3 (302C), associating attendees with different locations (or seats) in a meeting session can be referred to as seat tagging.

[0082] Thus, a trained ML model for a particular attendee can recognize different patterns or fluctuations in the attendee's attention, such as whether the attendee's attention score increases or decreases during certain situations or categories of meeting sessions. Furthermore, a personalized ML model can be trained based on such cumulative data that indicates the learning journey (at least in the form of attention) of a particular attendee (such as a student) with respect to factors such as subject / topic, class / standard (age group), meeting session type, meeting session location, seat tagging, educator information, environmental conditions, or attendee profile. Each trained model can also indicate the behavioral characteristics (i.e., actions, gestures, and interactions) of each attendee by training on the attention scores of past meeting sessions attended by the respective attendee. Furthermore, multiple ML models 106 created by the disclosed system 102 can function as a ranking and rating library for an education or learning system that can indicate the real-time performance of attendees / educators and provide different attendee / educator / content recommendations to develop a robust and holistic education or learning system. The multiple ML models 106 may further be stored in memory 204 for application in real-time scenarios such as those described in FIG.

[0083] Figure 4 illustrates example operations for meeting session control based on attention determination and application of the ML model for a particular attendee trained in Figure 3, according to an embodiment of the present disclosure. Figure 4 is described with reference to elements in Figures 1, 2, and 3. Figure 4 illustrates a block diagram 400 illustrating example operations 402A-402F described herein. The example operations illustrated in block diagram 400 may begin at 402A and may be performed by any computer system, apparatus, or device, such as system 102 of Figure 1 or circuit 202 of Figure 2. While example operations associated with one or more blocks in block diagram 400 are illustrated as discrete blocks, they may be divided into additional blocks, combined into fewer blocks, or eliminated depending on the particular implementation.

[0084] At 402A, a data acquisition operation can be performed. In the data acquisition operation, the circuit 202 can be configured to receive a first set of images 404 of a first attendee 406 of the multiple attendees 116. The first attendee 406 can be associated with (or attending or having attended) a first meeting session. Specifically, the first attendee 406 can be attending a first meeting session that can be different from the multiple meeting sessions (i.e., a first ML model 402 can be trained for the first attendee based on this meeting session, e.g., as described in FIG. 3 ). As described above, the multiple ML models 106 for the multiple attendees 116 can be trained based on data related to the multiple meeting sessions (i.e., attention scores or activity related to behaviors / gestures / interactions), but not based on data related to the first meeting session. In one embodiment, the first set of images 404 can be of multiple attendees (including the first attendee) that have attended (or are currently attending) the first meeting session. The circuitry 202 of the system 102 may receive the first set of images 404 from the server 110 or directly from one or more image capture devices located at different locations in the first meeting session. If the first meeting session is an online meeting session, the one or more image capture devices may be integrated into electronic devices (such as laptops, computers, mobile phones, etc.) used by attendees to attend the first meeting session.

[0085] In another embodiment, the circuit 202 can be further configured to control the audio capture device 108 (i.e., an audio capture device located near each attendee in the first meeting session) to capture interactions 408 between the first attendee 406 and the educator of the first meeting session. In some embodiments, the captured interactions 408 can relate to content presented in the first meeting session, or questions posed by the educator (or by the first attendee 406), or responses made by the first attendee 406 (or educator). For example, the first attendee can ask one or more questions related to the content, and the educator can respond to the first attendee 406 with respect to the one or more questions asked, or vice versa.

[0086] In another embodiment, the circuit 202 can be configured to determine experience information associated with each educator of the first meeting session. The experience information can indicate at least one of experience, evaluation, achievements, or feedback associated with each educator of the first meeting session. The circuit 202 can be further configured to determine content information associated with content presented during the first meeting session. The circuit 202 can be further configured to retrieve profile information associated with the first attendee or other attendees in the first meeting session. The profile information can indicate the first attendee's 406 preferences or interests in topics or content presented or discussed in the first meeting session. In another embodiment, the circuit 202 can be further configured to determine environmental information associated with the geolocation of at least one of the educator or the first attendee of the first meeting session. Details of the experience information, content information, profile information, and environmental information are shown, for example, in 302F of FIG. 3 .

[0087] At 402B, an activity detection operation may be performed. In the activity detection operation, the circuit 202 may be configured to detect a first set of activities that the first attendee 406 may perform during the first meeting session. The first set of activities may be detected based on applying one or more NN models 210 to the received first set of images 404. The first set of activities may be detected over a first period (e.g., several seconds or minutes) of the first meeting session. For example, the first set of activities of the first attendee 406 may be detected over a "10" minute period during the first meeting session. Details regarding the detection of the first set of activities are shown, for example, in 302C of FIG. 3 . For example, as described at 302C of FIG. 3 , the first activity set of the first attendee 406 can be related to at least one of an action performed by the first attendee 406, a gesture performed by the first attendee 406, a head pose of the first attendee 406, a body posture of the first attendee 406, a lip movement of the first attendee 406, a gaze of the first attendee 406, or a facial expression of the first attendee 406. In an embodiment, the circuit 202 can generate a three-dimensional (3D) map of the corresponding first meeting session including the first attendee based on the received first image set 404 (e.g., as described at 302B of FIG. 3 ). The circuit 202 can further apply one or more neural network (NN) models 210 to the generated three-dimensional (3D) map of the first meeting session to detect the first activity set of the first attendee 406.

[0088] At 402C, an interaction duration determination operation may be performed. In the interaction duration determination operation, the circuit 202 may be configured to determine a second duration (e.g., in seconds or minutes) of the captured interaction 408 between the first attendee 406 and the educator of the first meeting session. The circuit 202 may be further configured to compare the determined second duration with a threshold duration. If the determined second duration is less than the threshold duration, control may proceed to 402E. In other words, in such a case, the circuit 202 may discard the captured interaction 408. For example, if the threshold duration is between 5 and 10 seconds (i.e., the minimum response time to a question) and the captured interaction 408 is less than that number of seconds (e.g., 2 seconds), the circuit 202 may ignore the captured interaction 408 because it is considered to be of short duration that may not be considered a correct response for measuring the attendee's attentiveness or may not indicate appropriate response quality to the question asked. Otherwise, control can proceed to 402D.

[0089] In some embodiments, the circuitry 202 can compare the determined second duration with a threshold duration to determine the relevance of the captured interaction 408 to the content presented in the first meeting session. For example, if one or more questions asked by the first attendee are not related to the content, the educator may simply ignore the one or more questions asked or provide a short response such as "this is not relevant" or "I've already taught about this." Therefore, in such a case, the captured second duration may be short and shorter than the threshold duration. On the other hand, if one or more questions asked by the first attendee are related to the content, the educator may provide a detailed response, and therefore the captured second duration may be longer than the threshold duration. Also, relevant questions asked by an attendee or relevant responses given by an attendee may indicate the attendee's attentiveness to some extent.

[0090] In one embodiment, the circuit 202 can be further configured to utilize voice recognition technology to identify the voices of the first attendee 406 and the educator in the captured interaction 408. Detailed implementations of the above-described voice recognition technology are believed to be well known to those skilled in the art, and therefore, a detailed description of the above-described voice recognition technology is omitted from this disclosure for the sake of brevity. In another embodiment, the circuit 202 can be further configured to determine a sequence of interactions and an amplitude of the interactions in the captured interaction 408 between the first attendee 406 and the educator. The system 102 can be configured to determine whether the captured interaction 408 of the first attendee 406 is with the educator or another attendee present in the first meeting session based on the sequence of interactions and the amplitude of the interactions.

[0091] At 402D, a keyword determination operation can be performed. In the keyword determination operation, the circuit 202 can be configured to determine one or more keywords in the captured interaction 408. In one embodiment, the circuit 202 can be configured to apply one or more NN models 210 to the captured interaction 408 to determine the one or more keywords. In another embodiment, the circuit 202 can be configured to generate a transcription of the captured interaction 408 to further determine the one or more keywords in the captured interaction 408. In such a scenario, the one or more NN models 210 can include at least one natural language processing (NLP) model that can be trained to determine the one or more keywords in the captured interaction 408. The determined one or more keywords can be related to content presented in the first meeting session. The circuit 202 can be configured to analyze the determined keywords in the captured interaction 408 to determine whether the keywords are related to the topic of the presented content or whether they are related to the question posed. Based on such analysis and determination, the circuitry 202 can determine the attentiveness of the first attendee 406 during the first meeting session.

[0092] At 402E, an attention score calculation operation may be performed. In the attention score calculation operation, the circuit 202 may be configured to calculate a first attention score associated with the first attendee 406. The first attention score may indicate an attention level of the first attendee 406 in the first meeting session and may be calculated over a first time period based on the detected first set of activities. To calculate the attention score, the system 102 may be configured to calculate a first behavioral score and a first interaction score for the first attendee 406 based on the detected first set of activities. The circuit 202 may further be configured to calculate the first attention score based on the calculated first behavioral score and the calculated first interaction score. In another embodiment, the circuit 202 may be configured to determine a first facial expression of the first attendee 406 and further calculate a first attention score based on the determined first facial expression. Details regarding the calculation of the attention score (similar to the first attention score) are shown, for example, in 302I of FIG. 3.

[0093] In some embodiments, the circuit 202 can calculate a first attention score further based on the detected first set of activities, the determined second duration, and the determined one or more keywords in the captured interactions 408. In another embodiment, the first attention score can be calculated based on experience information, content information, profile information, and environmental information. In some embodiments, the circuit 202 can be configured to calculate a first concentration score and a first interaction score associated with the first attendee 406 based on the detected first set of activities, and / or the determined second duration, and / or the determined one or more keywords, and / or the experience information, and / or the content information, and / or the profile information, and / or the environmental information, and / or the determined first facial expression. The circuit 202 can be configured to calculate the first attention score based on the calculated first concentration score and the calculated first interaction score. Details regarding the calculation of the first attention score are shown, for example, in FIG. 3 .

[0094] At 402F, a trained ML model application operation may be performed. In the trained ML model application operation, the circuit 202 may be configured to apply a first ML model 402 to the calculated first attention score. Specifically, the circuit 202 may be configured to apply a first ML model 402 of the multiple ML models 106 (which may be trained at 302J of FIG. 3 ) to the calculated first attention score. For example, as described in FIG. 3 , the multiple ML models 106 may be trained based on multiple attention scores of multiple attendees 116 associated with multiple meeting sessions of different meeting categories. The system 102 may store the multiple ML models 106 in the memory 204. The multiple attendees 116 may have attended multiple meeting sessions in the past, and the multiple ML models 106 are trained based on these meeting sessions.

[0095] In one embodiment, the circuit 202 can apply a first ML model 402 to the calculated first attention score and the meeting category of a first meeting session that the first attendee 406 has attended or is currently attending. The first ML model 402 can be pre-trained based solely on data related to the first attendee 406, such that the first ML model 402 can be personalized for the first attendee. The first ML model 402 can be trained based on historical cumulative data (i.e., a set of past attention scores for the first attendee 406 calculated for different previously attended meeting sessions, e.g., as described in 302J of FIG. 3). In one embodiment, the circuit 202 can be configured to compare the calculated first attention score to a first threshold attention score. The first threshold attention score can be a minimum attention score associated with the first attendee 406 and can be determined based on a set of past attention scores that can be associated with the first attendee 406. If the calculated first attention score is less than the first threshold attention score, control may proceed to 402G. Otherwise, control may proceed to the end.

[0096] Based on the set of past attention scores (from which the first ML model 402 was trained), the circuit 202 can determine a point in time when the first attendee 406's attention significantly declines (i.e., a point in time during attendance at the first meeting session as indicated by the first attendee's 406's calculated first attention score). In other words, based on the first ML model 402 (i.e., an ML model trained based on the set of past attention scores from the first attendee 406 and different meeting categories), the circuit 202 can determine or predict a point in time when the first attendee 406's calculated first attention score (i.e., a first attendee attending a first meeting session of a particular category) will decline. For example, the training data for the first ML model 402 indicates that the first attendee's 406's past attention score is low for certain meeting sessions (such as online sessions without an educator, high-intensity meeting sessions, when the meeting session duration exceeds one hour, when the attendee is sitting in a rear position, or when the content is theoretical). In such a case, if the first meeting session is of a particular category that is considered to have been trained with a low attention score within the first ML model 402, the circuit 202 can predict in advance that the first attention score of the first attendee 406 may decrease while attending the first meeting session and require appropriate action or recommendation to increase the first attention score of the first attendee 406 during the remaining time of the first meeting session.

[0097] At 402G, a recommendation determination operation may be performed. In the recommendation determination operation, the circuit 202 may be configured to determine a first set of recommendations. The first set of recommendations may be associated with at least one of a first attendee 406, an educator of the first meeting session, or first content that may be presented in the first meeting session. The circuit 202 may determine the first set of recommendations based on application of the first ML model 402 to the calculated first attention score and a meeting category of the first meeting session. In an embodiment, the circuit 202 may determine the first set of recommendations based on determining or predicting (using the trained first ML model 402) that, with respect to the meeting category of the first meeting session, the first attention score of the first attendee 406 in the first meeting session is likely to decline (i.e., fall below a first threshold attention score). For example, if the first meeting session includes theoretical content, is long, and is conducted by a low-rated educator, the first attention score may be low or may decline after a period of time. Thus, the circuit 202 can determine a first set of recommendations for either the first attendee 406, the educator of the first meeting session, or the content presented in the first meeting session to further avoid a decrease in the first attention score, which, if followed, can increase the attention score of the first attendee 406 for the remainder of the first meeting session.

[0098] The determined first set of recommendations may include at least one of a first recommendation associated with the first attendee 406, a second recommendation associated with an educator of the first meeting session, or a third recommendation associated with content to be presented in the first meeting session. For example, the first recommendation associated with the first attendee 406 may include, but is not limited to, one or more individual performance reports including a detailed analysis of attention level breakdown across content and educators, one or more areas of focus / improvement, a recommended seating location in the meeting session, a recommendation to take an energy drink (such as coffee) or a break, appropriate content or educators based on the attendee's interests, etc. Secondary recommendations related to the educator of the meeting session may include, but are not limited to, content modifications (e.g., adding practical or real-time examples), suggestions for similar content with high ratings or many shares / likes (i.e., web links), content delivery modifications, alerting the educator to a significant change in the first attendee's 406 attention level, suggestions for taking a break or short break in the first meeting session, suggestions for updating the duration of the meeting session due to content, suggestions for changing the session's time due to environmental conditions, ways to improve environmental conditions (such as improving lighting or acoustic conditions), suggestions referencing specific qualifications, and suggestions for improving teaching skills or ratings. Third recommendations related to content presented in the first meeting session may include, but are not limited to, content modifications to add other types of content (e.g., self-explanatory content), content modifications to add interactive content (such as photos, videos, diagrams), content modifications to remove redundant content, content modifications to remove complex content, content modifications to add one or more practical sessions to the current theoretical content, etc.

[0099] In some embodiments, the first set of recommendations can serve as one or more feedbacks (such as 360-degree feedback) for a particular attendee or a particular educator. In another embodiment, the one or more feedbacks can be two-way feedback between an educator and a particular attendee. In some embodiments, an authority (such as an educational institution's administrative / governing board, a school board, or a school / college principal) can determine recommendations for making specific decisions (such as, but not limited to, changing to an educator with higher ratings or experience, changing the content, adding a specific practical session to the curriculum, shortening the duration of a specific session, adding breaks between meeting sessions, adding a specific educational counselor to the instructional staff, updating the overall instructional plan, updating the physical conditions or logistics of a meeting session).

[0100] At 402H, an output operation may be performed. In the output operation, the circuit 202 may be configured to output the determined first set of recommendations. For example, the system 102 may control the display device 206A to render the first set of recommendations for the first attendee 406 or the educator. In some embodiments, the circuit 202 may generate as output a notification for the first attendee 406 or the educator of the first meeting session based on the calculated first attention score and application of the first ML model 402 to the meeting category of the first meeting session. The generated notification may include at least one of the first set of recommendations. The notification may be an upfront alert that may be provided to the first attendee 406 or the educator of the first meeting session based on the determination or prediction of a low attention score for the first attendee 406 in the first meeting session. The low attention score of the first attendee 406 can be predicted based on application of the first ML model 402 (i.e., a personalization model trained for the first attendee 406) to the meeting category of the first meeting session and an attention score (i.e., a first attention score) for one or more activities (i.e., behaviors, gestures, interactions, and other factors) performed by the first attendee 406 calculated during the first meeting session. The advance notification can timely alert the educator or first attendee 406 to take appropriate action (such as a first set of recommendations) to avoid further decline in the first attention score or to increase the first attendee's 406's first attention score over the remaining duration of the first meeting session. The first set of recommendations provided to the attendee or educator can further enhance the attendee's learning process. The circuit 202 can further transmit the generated notification to an electronic device that may be associated with the first attendee 406 or the educator of the first meeting session. The electronic device may include suitable logic, circuitry, and interfaces that can be configured to receive notifications from the system 102 .The electronic device can be further configured to render the received notification. Examples of electronic devices can include, but are not limited to, a cellular phone, a mobile phone, a computing device, a smart phone, a gaming device, a mainframe machine, a server, a computer workstation, and / or a consumer electronics (CE) device. Thus, the disclosed system 102 can effectively control the meeting session to further enhance the attendee's learning journey using the output of recommendations determined using the personalized ML model trained for the attendee based on a real-time evaluation of the attendee's attention score.

[0101] Figure 5 illustrates an exemplary scenario for generating a simulated view of multiple attendees associated with a meeting session, according to an embodiment of the present disclosure. The description of Figure 5 is provided with reference to elements of Figures 1, 2, 3, and 4. Figure 5 illustrates a first diagram. Figure 5 illustrates a system 102 that may include multiple ML models 106. Additionally, multiple images 502 of multiple attendees 504A-504N that may be present in a particular meeting session are illustrated.

[0102] The system 102 can control a plurality of image capture devices 104 (as shown in FIG. 1 ) to capture a plurality of images 502 of a plurality of attendees 504A-504N. The plurality of attendees 504A-504N can include a first attendee 504A, a second attendee 504B, a third attendee 504C, and an Nth attendee 504N. If the meeting session is held at a physical location, the plurality of image capture devices 104 can include CCTV cameras that can be installed at specific locations within the physical space. If the meeting session is held virtually (i.e., in online mode), the plurality of image capture devices 104 can include at least one webcam associated with an electronic device through which the plurality of attendees 504A-504N can attend the meeting session.

[0103] The system 102 may receive a plurality of images 502 of a plurality of attendees 504A-504N from a plurality of image capture devices 104. The system 102 may be configured to identify each of the plurality of attendees 504A-504N based on receiving the plurality of images 502. In one embodiment, the system 102 may be configured to apply a facial recognition technique using one or more neural network models 210 to each of the plurality of images 502 to identify each of the plurality of attendees 504A-504N. Detailed implementations of the above-described facial recognition techniques are believed to be well known to those skilled in the art, and therefore, a detailed description of the above-described facial recognition techniques will be omitted from this disclosure for the sake of brevity.

[0104] If the meeting session is held at a physical location (such as a physical classroom), the system 102 can be configured to determine the location of each of the multiple attendees 504A-504N in the corresponding meeting session based on the received multiple images 502 (or based on a generated 3D map of the meeting session, e.g., as described in 302B of FIG. 3 ). In one embodiment, the circuit 202 of the system 102 can be configured to determine the location of each of the multiple attendees 504A-504N in the corresponding meeting session based on the captured depth information / multiple depth values. Based on the determined locations, the 3D map of the meeting session, and facial recognition of the attendees, the circuit 202 can tag or associate a particular attendee with a physical seat (such as at the back, front, center, or near a window / door) within the meeting session where the attendee is expected to be sitting while attending the meeting session.

[0105] If the meeting session is held virtually (i.e., a web conference or online session held over the Internet, etc.), the system 102 can be configured to generate a simulated view 506 of the corresponding meeting session. In the generated simulated view 506, each of the multiple attendees 504A-504N can be visualized as sitting next to each other (as in a physical class session). The generated simulated view 506 can be a computer-generated environment including realistic-looking scenes and objects that make each of the multiple attendees 504A-504N feel immersed in the physical meeting session and environment. In other words, the generated simulated view 506 can mimic a single, comprehensive view of the meeting session for each of the multiple attendees 504A-504N. In some embodiments, the circuit 202 can determine a particular online action (such as an online hand raise) taken by the first attendee 406 during the online meeting session and further translate this online action into a representation of a similar physical action (such as a physical hand raise). The circuitry 202 may further overlay this representation onto the corresponding attendee visualized in the simulated view 506 .

[0106] FIG. 6 illustrates an exemplary user interface for rendering dashboard information, according to an embodiment of the present disclosure. FIG. 6 is described with reference to elements in FIGS. 1, 2, 3, 4, and 5. FIG. 6 illustrates an electronic UI 600. The electronic UI 600 may be displayed on the display device 206A of the system 102 or on an electronic device associated with the educator or at least one of the plurality of attendees 116 based on receipt of a first user input. The first user input may be received through an application interface displayed on the display screen of the display device 206A or on a display screen of an electronic device associated with the educator or attendee. The application interface may be part of application software, such as, for example, a software development kit (SDK), a cloud server-based application, a web-based application, an OS-based application / application suite, an enterprise application, a mobile application, or the like.

[0107] In some embodiments, the system 102 can be configured to generate dashboard information. The generated dashboard information can be associated with a first attendee 406 of the plurality of attendees 116 and can be generated based on calculated attention scores for a series of meeting sessions that the first attendee 406 can attend. The generated dashboard can include one or more statistics for at least the series of meeting sessions. The circuit 202 can be further configured to render the generated dashboard information on the electronic UI 600. In some embodiments, dashboard information can be generated for multiple meeting sessions attended by a particular attendee or multiple attendees 116, for example, in a particular past week, month, or year.

[0108] The electronic UI 600 shows a series of UI elements, such as a first UI element 602, a second UI element 604, a third UI element 606, a fourth UI element 608, a fifth UI element 610, a sixth UI element 612, a seventh UI element 614, and an eighth UI element. The first UI element 602 may be labeled as “attendance status” and may indicate the presence / absence ratio of a plurality of attendees 116 in a first meeting session of the series of meeting sessions. In one example, the “attendance status” may indicate the presence / absence ratio of a first attendee 406 who previously attended the series of meeting sessions. The second UI element 604 may be labeled as “most attentive attendee” and may indicate at least one attendee who may have the highest attention score among the plurality of attendees 116 who attended the series of meeting sessions. The third UI element 606 may be labeled as “attendee who asked the most questions” and may indicate at least one attendee who was able to ask the greatest number of questions in the series of meeting sessions. The attendees who asked the most questions can be determined based on the attendees' activities and interactions in past meeting sessions, for example, as described in FIG. 3. A fourth UI element 608 can be labeled as "attendees needing help" and can indicate at least one attendee who may have the lowest attention score among multiple attendees 116 who attended a series of meeting sessions. For example, as described in 402G of FIG. 4, the disclosed system 102 can suggest a different set of recommendations to such attendees. In some embodiments, the disclosed system 102 can also suggest a prize or recognition from a meeting session educator for the "most attentive attendee" or the "attendee who asked the most questions." Doing so can create a sense of gamification among multiple attendees in a meeting session and further increase the attention score of each attendee in the meeting session.

[0109] The fifth UI element 610 may be labeled "Best Educators of the Last Month" and may show at least one educator present in past meeting sessions (e.g., occurring in the past month) in which a large number of attendees achieved a good attention score (or above a certain attention score threshold). The sixth UI element 612 may be labeled "Best Content (Topics) List" and may show at least one piece of content to which the first attendee 406 (or the largest number of attendees) paid more attention (i.e., had a high level of attention score). The seventh UI element 614 may be labeled "Attention Level Graph" and may show a graph (or statistic) between the attention level of the first attendee 406 and time for at least one meeting session of a series of meeting sessions. For example, the graph may show a decrease in a particular attendee's attention score based on the increase in time (or progress) of a particular meeting session. As another example, the statistics may show patterns or fluctuations in the attention scores of different attendees based on different factors related to the meeting session, such as the meeting session category, the meeting session duration, the educator, the content, the seating location, and the environmental conditions. Such graphs in the dashboard information for a particular attendee can further alert and / or encourage the attendee to effectively follow the recommendations in order to increase the attendee's attention score and improve their learning ability and related outcomes. Additionally, different statistics in the dashboard information (such as most attentive attendee) can encourage competition among attendees in a meeting session, thereby further improving the attention scores and learning ability of different attendees.

[0110] 6, it may be assumed that 70% of the plurality of attendees 116 are present in a first meeting session of a series of meeting sessions, while 30% of the plurality of attendees 116 are absent from the first meeting session. Attendees "A" and "B" may have the highest attention scores among the plurality of attendees 116. Attendees "C" and "D" may have asked the highest number of questions in the series of meeting sessions. Attendees "E" and "F" may have the lowest attention scores among the plurality of attendees 116. Educator "A" may be the best educator for the past month, and the list of best content (subjects) may include science, mathematics, English, computer science, and history.

[0111] In another embodiment, the generated dashboard information may include more statistics that may be related to the educator, the multiple attendees 116, or the content presented in the corresponding meeting session. The generated dashboard information may also include other statistics based on various parameters, such as content, grades, and years. In some embodiments, the generated dashboard information may be presented in a meeting session (e.g., a class) to encourage healthy competition among the multiple attendees 116. The generated dashboard information may further be used to evaluate the performance of a first attendee 406 and determine the first attendee's strengths and weaknesses. As shown in FIG. 6, the dashboard information may present different recommendation sets 616, including recommendations for the attendee (i.e., attendee 616A), recommendations for the educator (i.e., educator 616B), and recommendations for the content to be presented (i.e., content 616C), for example, as described in 402G of FIG. 4. The generated dashboard information may also be used by different educational institutions (or educators) to make appropriate decisions (such as following the recommendations) to effectively increase attendees' attention scores and improve their overall learning process. In some embodiments, the circuit 202 can determine attention scores not only for a particular attendee but also for a set of attendees present at a particular location in the meeting session. For example, the circuit 202 can determine attention scores for a set of attendees sitting in a particular row (such as the front row or the back row) in the meeting session. The circuit 202 can control the output of information related to the attention scores (i.e., the attention scores determined for a set of attendees) as dashboard information based on a request received from an attendee, an educator, or any educational / professional institution. Similarly, the dashboard information can present different information related to the attention scores in multiple ways, such as (but not limited to) by column, by row, by section, by age, by region, by educator, by year, and by a specified unit.Thus, the dashboard information generated by the disclosed system 102 serves as a real-time attention level meter for a particular meeting session (e.g., a lesson in progress), rendering the real-time attention scores of different attendees and further alerting the educator (using different user interface options such as highlighting or in other formats) to rising or falling attendee attention levels. Thus, the dashboard information can provide a real-time visualization of one or more statistics for at least a series of meeting sessions in terms of a gamification chart. Similarly, the dashboard information generated by the disclosed system 102 can indicate for which particular factors (such as a particular row / column / hotspot area, educator, content, meeting session type, session duration, subject / topic) attendee attention is good or needs to be improved with the real-time output of appropriate recommendations. Hotspot areas within a meeting session can correspond to areas where the attention scores of each attendee within that range are below a threshold attention score or above a threshold attention score.

[0112] In another embodiment, the disclosed system 102 can identify one or more malpractices that may be performed by a first attendee 406 of a first meeting session. The one or more malpractices may include, but are not limited to, cheating, talking with other attendees, illegally assisting an attendee by an educator (e.g., during an exam), using unauthorized materials in an exam venue, using illegal and abusive gestures by an attendee, and engaging in physical assault. In such a scenario, the circuit 202 can be configured to determine a pattern of a first set of detected activities performed by an attendee (e.g., the first attendee 406). Based on the determined pattern of the first set of activities, the circuit 202 can be configured to detect one or more malpractices during the first meeting session based on the determined pattern. In an embodiment, the circuit 202 can be configured to apply at least one of a plurality of ML models or one or more NN models 210 to the detected first set of activities to detect the malpractices. The system 102 can be configured to generate a first notification for the first attendee 406 or educator of the first meeting session based on the detected misconduct. In some embodiments, the system 102 can store information about the detected misconduct in memory 204 for future reference (such as for internal / external investigation or to take appropriate action to avoid such misconduct).

[0113] Figure 7 is a flowchart illustrating example operations for training an ML model for attention-based meeting session control according to an embodiment of the present disclosure. Figure 7 is described with reference to elements in Figures 1, 2, 3, 4, 5, and 6. Figure 7 illustrates a flowchart 700. Operations 702-710 can be performed on any computing device, such as system 102 or circuit 202. Operations can begin at 702 and proceed to 704.

[0114] At 704, a plurality of images 114 of a plurality of attendees 116 associated with a plurality of meeting sessions, including one or more meeting sessions with different meeting categories, may be received. In one or more embodiments, the circuitry 202 may be configured to receive a plurality of images 114 of a plurality of attendees 116 associated with a plurality of meeting sessions, including one or more meeting sessions with different meeting categories. Details regarding receiving the plurality of images 114 are shown, for example, in FIGS. 1, 3 (302A), and 4 (402A).

[0115] At 706, one or more activities performed by each of the plurality of attendees 116 during the corresponding meeting session can be detected over a period of time based on the received plurality of images 114. In one or more embodiments, the circuitry 202 can be configured to detect one or more activities performed by each of the plurality of attendees 116 during the corresponding meeting session over a period of time based on the received plurality of images 114. More details regarding detecting one or more activities are shown, for example, in FIGS. 1, 3 (302C), and 4 (402B).

[0116] At 708, an attention score for each of the plurality of attendees 116 may be calculated over a corresponding time period based on one or more detected activities associated with the corresponding attendee, the attention score indicating the attention level of each attendee in the corresponding meeting session. In one or more embodiments, the circuit 202 may be configured to calculate an attention score for each of the plurality of attendees 116 over a corresponding time period based on one or more detected activities associated with the corresponding attendee. Details regarding the calculation of attention scores are shown, for example, in FIG. 3 (302I).

[0117] At 710, an ML model can be trained for each of the plurality of attendees 116 based on the calculated attention score for each of the plurality of attendees 116 and the meeting category of the corresponding meeting session. In one or more embodiments, the circuit 202 can be configured to train an ML model for each of the plurality of attendees 116 based on the calculated attention score for each of the plurality of attendees 116 and the meeting category of the corresponding meeting session. Details regarding training the plurality of ML models 106 for the plurality of attendees 116 are shown, for example, in Figures 1 and 3 (302J). Control can proceed to an end.

[0118] Figure 8 is a flowchart illustrating example operations for meeting session control based on attention determination and application of ML models for particular attendees, according to an embodiment of the present disclosure. Figure 8 is described with reference to elements in Figures 1, 2, 3, 4, 5, 6, and 7. Figure 8 illustrates a flowchart 800. Operations 802-816 can be performed on any computing device, such as system 102 or circuit 202. Operations can begin at 802 and proceed to 804.

[0119] At 804, a plurality of ML models 106 can be stored that can be trained based on a plurality of attention scores of a plurality of attendees associated with a plurality of meeting sessions of different meeting categories. In one or more embodiments, the circuitry 202 can be configured to store a plurality of machine learning (ML) models that can be trained based on a plurality of attention scores of a plurality of attendees 116 associated with a plurality of meeting sessions of different meeting categories.

[0120] At 806, a first set of images 404 of a first attendee 406 associated with a first meeting session different from the plurality of meeting sessions of the plurality of attendees 116 may be received. In one or more embodiments, the circuit 202 may be configured to receive the first set of images 404 of a first attendee 406 associated with a first meeting session different from the plurality of meeting sessions, for example, as described in FIG. 4 (402A).

[0121] At 808, a first set of activities performed by the first attendee 406 during the first meeting session can be detected over a first time period based on the received first set of images 404. In one or more embodiments, the circuit 202 can be configured to detect a first set of activities performed by the first attendee 406 during the first meeting session over a first time period based on the received first set of images 404, for example, as described in FIG.

[0122] At 810, a first attention score indicative of an attention level of the first attendee 406 in the first meeting session associated with the first attendee 406 can be calculated over a first time period based on the detected first set of activities. In one or more embodiments, the circuit 202 can be configured to calculate a first attention score indicative of an attention level of the first attendee 406 in the first meeting session associated with the first attendee 406 over a first time period based on the detected first set of activities, for example, as described in FIG. 4 (402B).

[0123] At 812, a first machine learning (ML) model 402 of the plurality of ML models 106 may be applied to the calculated first attention score. In one or more embodiments, the circuit 202 may be configured to apply the first machine learning (ML) model 402 of the plurality of ML models 106 to the calculated first attention score, for example, as described in FIG. 4 (402F).

[0124] At 814, a first set of recommendations can be determined based on application of the first ML model 402 to the calculated first attention scores. In one or more embodiments, the circuit 202 can be configured to determine the first set of recommendations based on application of the first ML model 402 to the calculated first attention scores, for example, as described in FIG. 4 (402G).

[0125] At 816, the determined first set of recommendations may be output. In one or more embodiments, the circuit 202 may be configured to output the determined first set of recommendations, for example, as described in FIG. 4 (402H). Control may proceed to an end.

[0126] Various embodiments of the present disclosure may provide a non-transitory computer-readable medium and / or storage medium having stored thereon computer-executable instructions executable by a machine and / or computer, such as the system 102. The computer-executable instructions may cause the machine and / or computer to perform operations that may include receiving a plurality of images (e.g., a plurality of images 114) of a plurality of attendees (e.g., a plurality of attendees 116) associated with a plurality of meeting sessions. One or more of the plurality of meeting sessions may have different meeting categories. The operations may further include detecting, based on the received plurality of images 114, one or more activities that each of the plurality of attendees 116 may perform during the corresponding meeting session over a period of time. The operations may further include calculating an attention score for each of the plurality of attendees 116 over a corresponding period of time based on the detected one or more activities associated with the corresponding attendee. The attention score may indicate an attention level of each attendee in the corresponding meeting session. The operations may further include training a machine learning (ML) model (such as the first ML model 402) for each of the plurality of attendees 116 based on the calculated attention score of each of the plurality of attendees 116 and the meeting category of the corresponding meeting session.

[0127] Various embodiments of the present disclosure may provide a non-transitory computer-readable medium and / or storage medium having stored thereon computer-executable instructions executable by a machine and / or computer, such as system 102. The computer-executable instructions may cause the machine and / or computer to perform operations that may include storing a plurality of machine learning (ML) models (e.g., a plurality of ML models 106) that may be trained based on a plurality of attention scores of a plurality of attendees associated with a plurality of meeting sessions of different meeting categories. The operations may further include receiving a first set of images (e.g., first image set 404) of a first attendee (e.g., first attendee 406) of the plurality of attendees. The first attendee 406 may be associated with a first meeting session that may be different from the plurality of meeting sessions. The operations may further include detecting a first set of activities over a first time period that the first attendee 406 may perform during the first meeting session based on the received first image set 404. The operations may further include calculating a first attention score over a first time period, the first attention score being associated with the first attendee 406 based on the detected first set of activities. The first attention score may indicate an attention level of the first attendee 406 in the first meeting session. The operations may further include applying a first machine learning (ML) model 402 of the plurality of ML models 106 to the calculated first attention score. The operations may further include determining a first recommendation set based on application of the first ML model 402 to the calculated first attention score. The operations may further include outputting the determined first recommendation set.

[0128] An example aspect of the present disclosure may include a system (e.g., system 102 of FIG. 1 ) that may include a circuit (e.g., circuit 202). The circuit may be configured to receive a plurality of images (e.g., a plurality of images 114) of a plurality of attendees (e.g., a plurality of attendees 116) associated with a plurality of meeting sessions. One or more of the plurality of meeting sessions may have different meeting categories. The meeting category may correspond to at least one of a type of meeting session, a number of attendees in the meeting session, a duration of the meeting session, an average age of attendees in the meeting session, a topic of the meeting session, an experience of an educator in the meeting session, or content presented in the meeting session. The detected one or more activities performed by each of the plurality of attendees may be related to at least one of an action performed by the attendee, a gesture performed by the attendee, a head posture of the attendee, a body position of the attendee, lip movement of the attendee, a gaze of the attendee, or a facial expression of the attendee.

[0129] The circuitry 202 can be further configured to generate a first three-dimensional (3D) map of the corresponding meeting session including at least one of the plurality of attendees based on the received plurality of images. The circuitry 202 can be further configured to detect, over a period of time, one or more activities performed by each of the plurality of attendees during the corresponding meeting session based on the generated first 3D map.

[0130] According to an embodiment, the circuitry 202 may be configured to apply one or more neural network (NN) models (such as one or more NN models 210) to the received plurality of images 114. The circuitry 202 may be further configured to detect one or more activities performed by each of the plurality of attendees 116 based on the application of the one or more NN models 210.

[0131] According to an embodiment, the circuitry 202 may be further configured to calculate a concentration score for each of the plurality of attendees 116 based on the detected one or more activities associated with the corresponding attendee. The circuitry 202 may be further configured to calculate an interaction score for each of the plurality of attendees 116 based on the detected one or more activities associated with the corresponding attendee. The circuitry 202 may be further configured to calculate an attention score for each of the plurality of attendees 116 based on the calculated concentration score and the calculated interaction score of the corresponding attendee.

[0132] According to one embodiment, the circuit 202 may be configured to determine a facial expression of each of the plurality of attendees 116 based on the received plurality of images 114, and further calculate an attention score for each of the plurality of attendees 116 based further on the determined facial expression of the corresponding attendee, the calculated concentration score, and the calculated interaction score.

[0133] According to an embodiment, the circuit 202 may be configured to determine experience information associated with each educator of a corresponding meeting session of the plurality of meeting sessions. The experience information may indicate at least one of experience, evaluation, achievement, or feedback associated with each educator of the corresponding meeting session. The circuit 202 may be further configured to determine content information associated with content presented during each of the plurality of meeting sessions. The circuit 202 may be further configured to retrieve profile information associated with each of the plurality of attendees 116. The profile information may indicate preferences or interests in topics or content associated with the corresponding meeting session. The circuit 202 may be further configured to determine environmental information associated with the geolocation of the educator of each meeting session or at least one of the plurality of attendees 116. The circuit 202 may be further configured to calculate an attention score for each of the plurality of attendees 116 further based on the determined facial expression, the determined experience information, the determined content information, the retrieved profile information, the determined environmental information, the calculated concentration score, and the calculated interaction score for the corresponding attendee.

[0134] According to one embodiment, the circuitry 202 may be further configured to apply one or more neural network (NN) models to the received plurality of images. Based on the application, the circuitry 202 may be further configured to detect one or more objects associated with the plurality of attendees 116 in the received plurality of images 114. The circuitry 202 may be further configured to calculate an attention score associated with each of the plurality of attendees 116 based on the detected one or more objects.

[0135] According to an embodiment, the circuitry 202 may be configured to determine a first duration of at least one activity of the one or more activities performed by each of the plurality of attendees 116 during the corresponding meeting session, and further calculate an attention score for each of the plurality of attendees 116 based on the determined first duration of the at least one activity of the one or more activities.

[0136] According to an embodiment, the circuitry 202 may be further configured to generate dashboard information related to a first attendee 406 of the plurality of attendees 116 based on the calculated attention score for the series of meeting sessions in which the first attendee is present. The generated dashboard may include one or more statistics for at least the series of meeting sessions. The circuitry 202 may be further configured to output the generated dashboard information on a display device.

[0137] According to an embodiment of the present invention, the circuitry 202 may be configured to determine a position of each of the plurality of attendees 116 in the corresponding meeting session based on the received plurality of images 114, and further train an ML model for each of the plurality of attendees based on the calculated attention score and the determined position of each of the plurality of attendees 116 in the corresponding meeting session.

[0138] An example aspect of the present disclosure may include a system (such as system 102 of FIG. 1 ) that may include a memory (such as memory 204) that may be configured to store multiple machine learning (ML) models (such as multiple ML models 106) that may be trained based on multiple attention scores of multiple attendees (such as multiple attendees 116) associated with multiple meeting sessions of different meeting categories. System 102 may further include circuitry (such as circuit 202) that may be configured to receive a first set of images (such as first set of images 404) of a first attendee (such as first attendee 406) of the multiple attendees 116. The first attendee 406 may be associated with a first meeting session that may be different from the multiple meeting sessions. Circuit 202 may be further configured to detect a first set of activities performed by first attendee 406 during the first meeting session over a first time period based on the received first set of images 404. The circuit 202 may be further configured to calculate, based on the detected first set of activities, a first attention score associated with the first attendee 406 over a first time period, the first attention score indicative of the first attendee 406's attention level in the first meeting session. The circuit 202 may be further configured to apply a first machine learning (ML) model (such as the first ML model 402) of the plurality of ML models 106 to the calculated first attention score. The circuit 202 may be further configured to determine a first recommendation set based on application of the first ML model 402 to the calculated first attention score, and to output the determined first recommendation set.

[0139] According to an embodiment, the determined first set of recommendations includes at least one of a first recommendation related to a first attendee 406 of the plurality of attendees 116, a second recommendation related to an educator of the first meeting session, or a third recommendation related to content presented in the first meeting session.

[0140] According to an embodiment, the circuit 202 can be configured to compare the first attention score to a first threshold attention score. The circuit 202 can be further configured to determine a first set of recommendations based on the comparison. The circuit 202 can be further configured to generate a notification including at least one of the determined first set of recommendations.

[0141] According to an embodiment, the circuit 202 can be configured to control an audio capture device (e.g., audio capture device 108) to capture an interaction (e.g., interaction 408) between a first attendee 406 and an educator of the first meeting session during the first meeting session. The circuit 202 can be further configured to determine a second duration of the captured interaction 408. The circuit 202 can be further configured to determine, based on the captured interaction 408, one or more keywords in the captured interaction 408, and calculate a first attention score associated with the first attendee 406 based on the determined second duration and the determined one or more keywords in the captured interaction 408.

[0142] According to an embodiment, the circuitry 202 can be configured to generate a notification for the first attendee 406 or the educator of the first meeting session based on application of the first ML model 402 to the calculated first attention score and the meeting category of the first meeting session, the notification including at least one of the first set of recommendations.

[0143] According to an embodiment, the circuitry 202 can be configured to determine a pattern of the detected first set of activities performed by the first attendee 406. The circuitry 202 can be further configured to detect misconduct during the first meeting session based on the determined pattern, and generate a notification for the first attendee 406 or an educator of the first meeting session based on the detected misconduct.

[0144] The present disclosure can be implemented in hardware or a combination of hardware and software. The present disclosure can be implemented in a centralized manner in at least one computer system, or in a distributed manner where different elements can be distributed across several interconnected computer systems. Any computer system or other device adapted to perform the methods described herein can be suitable. The combination of hardware and software can be a general-purpose computer system that includes a computer program that, when loaded and executed, can control the computer system to perform the methods described herein. The present disclosure can be implemented in hardware, including portions of integrated circuits that also perform other functions.

[0145] The present disclosure may also be embodied in a computer program product, which includes all features that enable the implementation of the methods described herein and which is capable of executing these methods when loaded into a computer system. A computer program in this context means any expression, in any language, code or notation, of a set of instructions intended to cause a system having information processing capabilities to perform a particular function, either directly, or after a) conversion into another language, code or notation, or b) reproduction in a different content form, or both.

[0146] While the present disclosure has been described with reference to several embodiments, those skilled in the art will recognize that various modifications may be made and equivalents may be substituted without departing from the scope of the disclosure. Additionally, many modifications may be made to adapt a particular situation or material to the teachings of the disclosure without departing from the scope of the disclosure. Therefore, it is not intended that the disclosure be limited to the particular embodiments disclosed, but rather, it is intended to include all embodiments falling within the scope of the appended claims. [Explanation of symbols]

[0147] 104 Multiple Image Capture Devices 106 Multiple ML Models 114 Multiple Images 210 one or more neural network models 300 Block Diagram 302A Data Acquisition 302B 3D map generation 302C Activity Detection 302D Concentration score calculation 302E Interaction score calculation 302F Information Decision 302G facial expression determination 302H Object Detection 302I Attention score calculation 302J ML model training 304 Multiplexer

Claims

1. receiving a plurality of images of a plurality of attendees associated with a plurality of meeting sessions, the plurality of meeting sessions including one or more meeting sessions of different meeting categories; detecting one or more activities performed by each of the plurality of attendees during a corresponding meeting session over a period of time based on the plurality of received images; calculating an attention score for each of the plurality of attendees over a corresponding time period based on the detected one or more activities associated with the corresponding attendee, the attention score being indicative of each attendee's level of attention in the corresponding meeting session; training a machine learning (ML) model for each of the plurality of attendees based on the calculated attention score for each of the plurality of attendees and the meeting category of the corresponding meeting session. The circuit is configured to A system characterized by:

2. the meeting category corresponds to at least one of a type of the meeting session, a number of attendees of the meeting session, a duration of the meeting session, an average age of attendees of the meeting session, a topic of the meeting session, an educator experience of the meeting session, or content presented in the meeting session; The system of claim 1 .

3. the detected one or more activities performed by each of the plurality of attendees relate to at least one of an action performed by the attendee, a gesture performed by the attendee, a head pose of the attendee, a body position of the attendee, lip movement of the attendee, a gaze of the attendee, or a facial expression of the attendee. The system of claim 1 .

4. The circuit comprises: generating a first three-dimensional (3D) map of the corresponding meeting session including at least one of the plurality of attendees based on the received plurality of images; detecting the one or more activities performed by the plurality of attendees based on the generated first 3D map; The system of claim 1 further configured to:

5. The circuit comprises: calculating a focus score for each of the plurality of attendees based on the detected one or more activities associated with the corresponding attendee; calculating an interaction score for each of the plurality of attendees based on the detected one or more activities associated with the corresponding attendee; calculating an attention score for each of the plurality of attendees based on the calculated focus score and the calculated interaction score of the corresponding attendee; The system of claim 1 further configured to:

6. The circuit comprises: determining a facial expression of each of the plurality of attendees based on the plurality of received images; calculating an attention score for each of the plurality of attendees further based on the determined facial expression of the corresponding attendee, the calculated concentration score, and the calculated interaction score. The system of claim 5 further configured to:

7. The circuit comprises: determining experience information associated with each educator for a corresponding meeting session of the plurality of meeting sessions and indicative of at least one of an experience, an evaluation, an achievement, or feedback associated with each educator for the corresponding meeting session; determining content information relating to content presented during each of the plurality of meeting sessions; calculating an attention score for each of the plurality of attendees further based on the determined experience information, the determined content information, the calculated focus score and the calculated interaction score for the corresponding attendee. The system of claim 5 further configured to:

8. The circuit comprises: retrieving profile information associated with each of the plurality of attendees indicative of preferences or interests in topics or content related to the corresponding meeting session; calculating the attention score for each of the plurality of attendees further based on the retrieved profile information of the corresponding attendee, the calculated focus score, and the calculated interaction score. The system of claim 5 further configured to:

9. The circuit comprises: determining environmental information associated with a geolocation of an educator or at least one of the plurality of attendees for each meeting session; calculating an attention score for each of the plurality of attendees further based on the determined environmental information, the calculated concentration score, and the calculated interaction score of the corresponding attendee. The system of claim 5 further configured to:

10. The circuit comprises: applying one or more neural network (NN) models to the received images; detecting the one or more activities performed by each of the plurality of attendees based on application of the one or more neural network models; calculating an attention score for each of the plurality of attendees over the period based on the detected one or more activities associated with the corresponding attendee; The system of claim 1 further configured to:

11. The circuit comprises: applying one or more neural network (NN) models to the received images; Detecting one or more objects associated with the plurality of attendees in the plurality of received images based on the application; calculating an attention score associated with each of the plurality of attendees based on the detected one or more objects; The system of claim 1 further configured to:

12. The circuit comprises: determining a first duration of at least one activity of the one or more activities performed by each of the plurality of attendees during the corresponding meeting session; calculating an attention score for each of the plurality of attendees based on the determined first duration of the at least one activity of the one or more activities; The system of claim 1 further configured to:

13. The circuit comprises: receiving a first set of images of a first attendee of the plurality of attendees associated with a first meeting session different from the plurality of meeting sessions; detect a first set of activities performed by the first attendee during the first meeting session over a first time period based on the received first set of images; calculating a first attention score associated with the first attendee for the first time period based on the detected first set of activities; applying a first ML model of a plurality of trained machine learning (ML) models to the calculated first attention score, the first ML model being trained based on a set of historical attention scores associated with the first attendee; determining a first set of recommendations based on application of the first ML model to the calculated first attention scores; outputting the determined first set of recommendations; The system of claim 1 further configured to:

14. the circuitry is further configured to generate a notification for the first attendee or an educator of the first meeting session based on application of the first ML model to the calculated first attention score and a meeting category of the first meeting session; the notification includes at least one of the first set of recommendations. The system of claim 13.

15. The circuit comprises: generating dashboard information including one or more statistics for at least the set of meeting sessions associated with a first attendee of the plurality of attendees based on the calculated attention scores for the set of meeting sessions attended by the first attendee; outputting the generated dashboard information on a display device; The system of claim 1 further configured to:

16. The circuit comprises: determining a location of each of the plurality of attendees in the corresponding meeting session based on the plurality of received images; training the ML model for each of the plurality of attendees based on the calculated attention score and the determined position of each of the plurality of attendees in the corresponding meeting session. The system of claim 1 further configured to:

17. a memory configured to store a plurality of machine learning (ML) models trained based on a plurality of attention scores of a plurality of attendees associated with a plurality of meeting sessions of different meeting categories; a circuit communicatively coupled to the memory; The circuit comprises: receiving a first set of images of a first attendee of the plurality of attendees associated with a first meeting session different from the plurality of meeting sessions; detect a first set of activities performed by the first attendee during the first meeting session over a first time period based on the received first set of images; calculating a first attention score associated with the first attendee based on the detected first set of activities over the first time period, the first attention score being indicative of an attention level of the first attendee in the first meeting session; applying a first machine learning (ML) model of a plurality of ML models to the first calculated attention score; determining a first set of recommendations based on application of the first ML model to the calculated first attention scores; outputting the determined first set of recommendations; It is configured as follows: A system characterized by:

18. the determined first set of recommendations includes at least one of a first recommendation associated with a first attendee of the plurality of attendees, a second recommendation associated with an educator of the first meeting session, or a third recommendation associated with content presented at the first meeting session.

20. The system of claim 17.

19. The circuit comprises: comparing the first attention score to a first threshold attention score; determining a first set of recommendations based on the comparison; generating a notification including at least one of the determined first set of recommendations; The system of claim 17 further configured to:

20. The circuit comprises: controlling an audio capture device during the first meeting session to capture an interaction between the first attendee and an educator of the first meeting session; determining a second duration of the captured interaction; determining one or more keywords in the captured interactions based on the captured interactions; calculating the first attention score associated with the first attendee further based on the determined second duration and the determined one or more keywords in the captured interaction; The system of claim 17 further configured to:

21. the circuitry is further configured to generate a notification for the first attendee or an educator of the first meeting session based on application of the first ML model to the calculated first attention score and a meeting category of the first meeting session; the notification includes at least one of the first set of recommendations.

20. The system of claim 17.

22. The circuit comprises: determining a pattern of the detected first set of activities performed by the first attendee; Detecting illegal activity during the first meeting session based on the determined pattern; generating a notification for the first attendee or the educator of the first meeting session based on the detected misconduct; The system of claim 17 further configured to:

23. In the system, receiving a plurality of images of a plurality of attendees associated with a plurality of meeting sessions, the meeting sessions including one or more meeting sessions of different meeting categories; detecting, based on the received images, one or more activities performed by each of the plurality of attendees during a corresponding meeting session over a period of time; calculating an attention score for each of the plurality of attendees for a corresponding time period based on the detected one or more activities associated with the corresponding attendee, the attention score being indicative of a level of attention of each attendee in the corresponding meeting session; training a machine learning (ML) model for each of the plurality of attendees based on the calculated attention score for each of the plurality of attendees and the meeting category of the corresponding meeting session; The method according to claim 1, further comprising:

Images