An active AI intelligent teaching feedback system based on classroom behavior data analysis

Abstract

This invention provides a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis, comprising: a data acquisition layer, a class brain, a proactive feedback engine, a campus super brain, and an application service layer. The data acquisition layer, configured in the classroom, includes a high-definition camera, an omnidirectional array microphone, a smart blackboard, and behavior sensors for real-time acquisition of multimodal classroom data. The class brain, an edge computing node deployed in the classroom, communicates with the data acquisition layer and incorporates a multimodal AI model. This proactive AI-powered intelligent teaching feedback system, through its two-tiered architecture of "class brain and campus super brain," achieves localized real-time processing and proactive feedback of multimodal classroom data. The class brain, as an edge computing node deployed in the classroom, incorporates a 20 TOPS high-performance NPU, enabling real-time behavior recognition and feedback judgment at the data source, fundamentally solving data security risk issues.

Description

Technical Field

[0001] This invention relates to the intersection of artificial intelligence and smart education, and in particular to a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis. Background Technology

[0002] With the rapid development of artificial intelligence technology, AI is being applied more and more widely in the field of education. National and local governments have successively introduced relevant policies on "artificial intelligence + education", clearly proposing to deeply integrate AI into all elements and processes of education and teaching, build high-quality, independent and controllable educational datasets, and create large-scale models specifically for education.

[0003] However, schools currently face three major pain points in the process of implementing AI applications:

[0004] First, there is insufficient computing power support. The application of AI in education is growing explosively. Scenarios such as real-time analysis of classroom audio and video and multimodal teaching assessment have increased the demand for computing power exponentially. However, most schools lack local computing servers, and traditional server architectures are difficult to expand as needed, resulting in high latency and high cost of real-time AI analysis, which seriously affects the teaching experience.

[0005] Second, data is scattered and isolated. Data such as academic affairs system, personnel system, classroom videos, and scanned homework are stored in different systems with inconsistent formats, forming data silos and making it impossible to effectively train education-specific models.

[0006] Third, data security risks. Over-reliance on public cloud big data models on the Internet leads to the risk of sensitive data such as students' names, grades, and behavioral patterns being leaked. Some big data model service providers have servers located overseas, making it difficult for schools to control data sovereignty. There have been cases of schools being summoned for talks due to data leaks.

[0007] In terms of traditional teaching feedback, most existing classroom analysis systems are post-event statistical reports, lacking real-time and proactive teaching intervention capabilities. Teachers can only review the classroom after class and cannot obtain immediate feedback during the teaching process, making it difficult to adjust teaching strategies in a timely manner. In addition, existing systems often require uploading classroom audio and video to the cloud for processing, posing significant privacy and security risks.

[0008] Therefore, it is necessary to provide a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis to solve the aforementioned technical problems. Summary of the Invention

[0009] This invention provides a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis, which solves the problems of delayed classroom feedback, insufficient computing power, data silos, and data security risks in existing technologies.

[0010] To address the aforementioned technical problems, this invention provides a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis, comprising:

[0011] Data acquisition layer, class brain, proactive feedback engine, campus super brain, and application service layer;

[0012] The data acquisition layer, configured in the classroom, includes a high-definition camera, an omnidirectional array microphone, a smart blackboard, and behavior sensors, used to collect multimodal classroom data in real time;

[0013] The class brain is an edge computing node deployed on the classroom side, which communicates with the data acquisition layer. It has a built-in multimodal AI model for real-time reasoning of the multimodal data, recognizing student behavior, teacher teaching behavior, voice content and attention state, and generating classroom behavior analysis results.

[0014] The active feedback engine is integrated with the class brain and has a built-in rule engine. It is used to determine whether to trigger active feedback based on preset attention thresholds, interaction thresholds, emotion thresholds or teaching behavior thresholds, and to send feedback instructions to the teacher terminal and / or classroom equipment when triggered.

[0015] The campus super brain is a school-level central server deployed in the school's computer room. It is connected to the class brain and is used to receive and aggregate desensitized behavioral data uploaded by the class brain, perform model training and optimization, and distribute the updated model and intelligent applications to the class brain.

[0016] The application service layer, which communicates with the campus super brain, includes a classroom analysis report system, a teaching and research support system, a student profiling system, and a teaching improvement suggestion system, which are used to provide visualization services to teachers and administrators.

[0017] Preferably, the multimodal data includes video data, audio data, whiteboard data, and behavioral trajectory data.

[0018] Preferably, the class brain includes: a behavior recognition module, a speech-to-text module, an attention analysis module, and a teaching behavior analysis module;

[0019] The behavior recognition module is used to identify students' focus, interactive behavior, and emotional state.

[0020] The speech-to-text module is used to transcribe the teacher's speech into text in real time and generate a teaching transcript;

[0021] The attention analysis module is used to analyze the overall attention curve in the classroom and individual attention fluctuations;

[0022] The teaching behavior analysis module is used to analyze teachers' questioning frequency, speaking speed, movement, and classroom management behavior.

[0023] Preferably, the active feedback engine includes: a rule configuration unit, a teacher terminal push unit, and a classroom equipment control unit;

[0024] The rule configuration unit allows teachers or administrators to customize feedback trigger conditions and feedback content;

[0025] The teacher terminal push unit is used to send prompt information to the teacher's mobile phone, smart blackboard or tablet computer;

[0026] The classroom equipment control unit is used to trigger changes in the classroom lights, prompts, or screen displays.

[0027] Preferably, the proactive feedback includes: when the students' overall attention is below a preset threshold, sending a prompt to the teacher that "students' attention has decreased, interaction is recommended";

[0028] When a prolonged period of no response is detected, a prompt is sent to the teacher suggesting that they adjust their questioning style or pace.

[0029] When a teacher speaks too fast, a notification will be sent to the teacher saying "The speaking speed is too fast, it is recommended to slow down";

[0030] When active student participation is detected, a positive incentive signal is sent to the classroom equipment.

[0031] Preferably, the classroom brain uses a domestically produced edge computing chip with a built-in tri-core NPU, providing 20 TOPS of AI computing power and supporting real-time decoding of 32 channels of 1080P video streams, enabling localized processing of classroom data and ensuring that sensitive teaching data does not leave the classroom.

[0032] Preferably, the campus super brain adopts a domestically developed information technology innovation architecture, including the Kunpeng processor and the Ascend AI accelerator card, and has a built-in controllable data set hub, a campus business application hub, and a one-stop intelligent agent hub, supporting local fine-tuning and training of education-specific models.

[0033] Preferably, it also includes a full-process visualized teaching and research robot, which supports administrators or teaching and research groups to remotely listen to classes, observe teaching methods and student participation, and provide multi-dimensional AI analysis reports.

[0034] Preferably, it also includes an audio and video automatic segmentation robot, which is used to automatically segment recorded videos according to classroom knowledge points and generate short videos of knowledge points for students to review and teachers to use for teaching and research.

[0035] Preferably, a closed-loop data transfer mechanism is adopted between the class brain and the campus super brain: the class brain collects and preprocesses data, uploads it to the campus super brain for model training, the campus super brain distributes the updated model, the class brain applies the model to generate feedback, and the feedback effect data is transmitted back to the campus super brain, forming a continuous optimization closed loop.

[0036] Compared with related technologies, the proactive AI intelligent teaching feedback system based on classroom behavior data analysis provided by this invention has the following beneficial effects:

[0037] This invention provides a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis. Through a two-tiered architecture of "Classroom Brain and Campus Super Brain," it achieves localized real-time processing and proactive feedback of multimodal classroom data. The Classroom Brain, as an edge computing node deployed in the classroom, is equipped with a built-in 20 TOPS high-performance NPU, enabling real-time behavior recognition and feedback judgment at the data source. This eliminates the need to upload sensitive data to the cloud, fundamentally solving data security risks. The proactive feedback engine, based on a rule engine, proactively pushes teaching suggestions to teachers or controls classroom equipment during class, realizing a paradigm shift in teaching support from "post-event statistics" to "in-event intervention." The Campus Super Brain, as the school-level center, aggregates anonymized data from each class for model training and optimization, forming a continuously evolving educational model. This model is then distributed to the Classroom Brain through a closed-loop mechanism, giving the system self-learning and self-optimization capabilities. This invention effectively addresses three major pain points: insufficient school computing power, data silos, and security risks, significantly improving classroom teaching quality and teachers' teaching abilities. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of a preferred embodiment of an active AI intelligent teaching feedback system based on classroom behavior data analysis provided by the present invention.

[0039] Figure 2 A schematic diagram of an active feedback closed-loop process;

[0040] Figure 3 for Figure 1 The diagram shows a proactive AI-powered intelligent teaching feedback system. Detailed Implementation

[0041] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0042] Please refer to the following: Figure 1 , Figure 2 and Figure 3 ,in, Figure 1 This is a schematic diagram of a preferred embodiment of an active AI intelligent teaching feedback system based on classroom behavior data analysis provided by the present invention. Figure 2 A schematic diagram of an active feedback closed-loop process; Figure 3 for Figure 1 The diagram shows a proactive AI-powered intelligent teaching feedback system. A proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis includes:

[0043] Data acquisition layer, class brain, proactive feedback engine, campus super brain, and application service layer;

[0044] The data acquisition layer, configured in the classroom, includes a high-definition camera, an omnidirectional array microphone, a smart blackboard, and behavior sensors, used to collect multimodal classroom data in real time;

[0045] The class brain is an edge computing node deployed on the classroom side, which communicates with the data acquisition layer. It has a built-in multimodal AI model for real-time reasoning of the multimodal data, recognizing student behavior, teacher teaching behavior, voice content and attention state, and generating classroom behavior analysis results.

[0046] The active feedback engine is integrated with the class brain and has a built-in rule engine. It is used to determine whether to trigger active feedback based on preset attention thresholds, interaction thresholds, emotion thresholds or teaching behavior thresholds, and to send feedback instructions to the teacher terminal and / or classroom equipment when triggered.

[0047] The campus super brain is a school-level central server deployed in the school's computer room. It is connected to the class brain and is used to receive and aggregate desensitized behavioral data uploaded by the class brain, perform model training and optimization, and distribute the updated model and intelligent applications to the class brain.

[0048] The application service layer, which communicates with the campus super brain, includes a classroom analysis report system, a teaching and research support system, a student profiling system, and a teaching improvement suggestion system, which are used to provide visualization services to teachers and administrators.

[0049] The multimodal data includes video data, audio data, whiteboard data, and behavioral trajectory data.

[0050] The class brain includes: a behavior recognition module, a speech-to-text module, an attention analysis module, and a teaching behavior analysis module;

[0051] The behavior recognition module is used to identify students' focus, interactive behavior, and emotional state.

[0052] The speech-to-text module is used to transcribe the teacher's speech into text in real time and generate a teaching transcript;

[0053] The attention analysis module is used to analyze the overall attention curve in the classroom and individual attention fluctuations;

[0054] The teaching behavior analysis module is used to analyze teachers' questioning frequency, speaking speed, movement, and classroom management behavior.

[0055] The active feedback engine includes: a rule configuration unit, a teacher terminal push unit, and a classroom equipment control unit;

[0056] The rule configuration unit allows teachers or administrators to customize feedback trigger conditions and feedback content;

[0057] The teacher terminal push unit is used to send prompt information to the teacher's mobile phone, smart blackboard or tablet computer;

[0058] The classroom equipment control unit is used to trigger changes in the classroom lights, prompts, or screen displays.

[0059] The proactive feedback includes: when students' overall attention falls below a preset threshold, sending a prompt to the teacher that "students' attention has decreased, interaction is recommended";

[0060] When a prolonged period of no response is detected, a prompt is sent to the teacher suggesting that they adjust their questioning style or pace.

[0061] When a teacher speaks too fast, a notification will be sent to the teacher saying "The speaking speed is too fast, it is recommended to slow down";

[0062] When active student participation is detected, a positive incentive signal is sent to the classroom equipment.

[0063] The classroom brain uses a domestically produced edge computing chip with a built-in tri-core NPU, providing 20 TOPS of AI computing power and supporting real-time decoding of 32 channels of 1080P video streams, enabling localized processing of classroom data and ensuring that sensitive teaching data does not leave the classroom.

[0064] The campus super brain has a built-in controllable data set hub, a campus business application hub, and a one-stop intelligent agent hub. It is used to receive and aggregate desensitized behavioral data uploaded by the class brain, perform local fine-tuning and training of education-specific models, and distribute the updated models and intelligent applications to the class brain.

[0065] The campus superbrain adopts a domestically developed information technology architecture, including the Kunpeng processor and the Ascend AI accelerator card. It has a built-in controllable data set hub, a campus business application hub, and a one-stop intelligent agent hub, and supports local fine-tuning and training of education-specific models.

[0066] It also includes a fully visualized teaching and research robot, which supports administrators or teaching and research groups to remotely listen to classes, observe teaching methods and student participation, and provide multi-dimensional AI analysis reports.

[0067] It also includes an automatic audio and video segmentation robot, which is used to automatically segment recorded videos according to classroom knowledge points and generate short videos of knowledge points for students to review and teachers to use for teaching and research.

[0068] The class brain and the campus super brain adopt a closed-loop data flow mechanism: the class brain collects and preprocesses data, uploads it to the campus super brain for model training, the campus super brain distributes the updated model, the class brain applies the model to generate feedback and the feedback effect data is sent back to the campus super brain, forming a continuous optimization closed loop.

[0069] The teaching improvement suggestion system provides teachers with personalized teaching improvement suggestions based on long-term data analysis. The application service layer provides teachers and administrators with a visual web interface or mobile application, supporting data querying, analysis and export.

[0070] The classroom analysis report system is used to generate classroom behavior analysis reports for single or multiple lessons; the teaching and research support system is used to support online lesson observation and teaching evaluation for teaching and research groups; and the student profiling system is used to construct learning behavior profiles for individual students and groups.

[0071] The active feedback engine is integrated with the class brain, as shown in the figure. It has a built-in rule engine, including: a rule configuration unit, which allows teachers or administrators to customize feedback trigger conditions (such as attention threshold below 60%, no response time exceeding 5 seconds, teacher speaking speed exceeding 180 words / minute, etc.) and feedback content through a visual interface; a teacher terminal push unit, which is used to send prompt information to the teacher's mobile phone, smart blackboard or tablet via Wi-Fi or Bluetooth; and a classroom equipment control unit, which is used to trigger changes in classroom lighting (such as flashing reminders), prompt sounds (such as gentle prompt sounds), or screen displays (such as displaying encouraging words) through IoT protocols.

[0072] The working principle of the proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis provided by this invention is as follows:

[0073] First, the high-definition cameras, omnidirectional array microphones, smart blackboards, and behavior sensors in the data acquisition layer collect multimodal data in the classroom in real time, including video, audio, blackboard writing, and behavioral trajectory data, and transmit them to the class brain.

[0074] Secondly, the class brain utilizes a built-in multimodal AI model to perform real-time reasoning on the data. The behavior recognition module identifies student focus, interactive behaviors, and emotional states; the speech-to-text module transcribes the teacher's speech into text in real time; the attention analysis module analyzes the overall classroom attention curve; and the teaching behavior analysis module analyzes the teacher's questioning frequency, speaking speed, and movement, generating classroom behavior analysis results.

[0075] Then, the proactive feedback engine determines whether to trigger proactive feedback based on the preset attention threshold, interaction threshold, etc. in the rule configuration unit. If the judgment result meets the triggering conditions, it sends a prompt message to the teacher's mobile phone or smart blackboard through the teacher terminal push unit, or controls the classroom lights, prompt sounds, or screen displays through the classroom equipment control unit.

[0076] Meanwhile, the class brain uploads the anonymized behavioral data to the campus super brain. The campus super brain aggregates data from all classes, performs data governance through a controllable data set hub, and trains and optimizes models through a one-stop intelligent agent hub, forming updated education-specific models and intelligent applications. These are then distributed to each class brain via the campus network, enabling the system to continuously evolve.

[0077] Finally, based on the data gathered by the campus super brain, the application service layer provides visualization services such as classroom analysis reports, teaching and research assistance, student profiles, and teaching improvement suggestions, forming a closed loop of teaching support from data collection and real-time feedback to long-term analysis.

[0078] Compared with related technologies, the proactive AI intelligent teaching feedback system based on classroom behavior data analysis provided by this invention has the following beneficial effects:

[0079] This invention provides a proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis. Through a two-tiered architecture of "Classroom Brain and Campus Super Brain," it achieves localized real-time processing and proactive feedback of multimodal classroom data. The Classroom Brain, as an edge computing node deployed in the classroom, is equipped with a built-in 20 TOPS high-performance NPU, enabling real-time behavior recognition and feedback judgment at the data source. This eliminates the need to upload sensitive data to the cloud, fundamentally solving data security risks. The proactive feedback engine, based on a rule engine, proactively pushes teaching suggestions to teachers or controls classroom equipment during class, realizing a paradigm shift in teaching support from "post-event statistics" to "in-event intervention." The Campus Super Brain, as the school-level center, aggregates anonymized data from each class for model training and optimization, forming a continuously evolving educational model. This model is then distributed to the Classroom Brain through a closed-loop mechanism, giving the system self-learning and self-optimization capabilities. This invention effectively addresses three major pain points: insufficient school computing power, data silos, and security risks, significantly improving classroom teaching quality and teachers' teaching abilities.

[0080] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis, characterized in that: include: Data acquisition layer, class brain, proactive feedback engine, campus super brain, and application service layer; The data acquisition layer, configured in the classroom, includes a high-definition camera, an omnidirectional array microphone, a smart blackboard, and behavior sensors, used to collect multimodal classroom data in real time; The class brain is an edge computing node deployed on the classroom side, which communicates with the data acquisition layer. It has a built-in multimodal AI model for real-time reasoning of the multimodal data, recognizing student behavior, teacher teaching behavior, voice content and attention state, and generating classroom behavior analysis results. The active feedback engine is integrated with the class brain and has a built-in rule engine. It is used to determine whether to trigger active feedback based on preset attention thresholds, interaction thresholds, emotion thresholds or teaching behavior thresholds, and to send feedback instructions to the teacher terminal and / or classroom equipment when triggered. The campus super brain is a school-level central server deployed in the school's computer room. It is connected to the class brain and is used to receive and aggregate desensitized behavioral data uploaded by the class brain, perform model training and optimization, and distribute the updated model and intelligent applications to the class brain. The application service layer, which communicates with the campus super brain, includes a classroom analysis report system, a teaching and research support system, a student profiling system, and a teaching improvement suggestion system, which are used to provide visualization services to teachers and administrators.

2. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The multimodal data includes video data, audio data, whiteboard data, and behavioral trajectory data.

3. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The class brain includes: a behavior recognition module, a speech-to-text module, an attention analysis module, and a teaching behavior analysis module; The behavior recognition module is used to identify students' focus, interactive behavior, and emotional state. The speech-to-text module is used to transcribe the teacher's speech into text in real time and generate a teaching transcript; The attention analysis module is used to analyze the overall attention curve in the classroom and individual attention fluctuations; The teaching behavior analysis module is used to analyze teachers' questioning frequency, speaking speed, movement, and classroom management behavior.

4. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The active feedback engine includes: a rule configuration unit, a teacher terminal push unit, and a classroom equipment control unit; The rule configuration unit allows teachers or administrators to customize feedback trigger conditions and feedback content; The teacher terminal push unit is used to send prompt information to the teacher's mobile phone, smart blackboard or tablet computer; The classroom equipment control unit is used to trigger changes in the classroom lights, prompts, or screen displays.

5. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The proactive feedback includes: when students' overall attention falls below a preset threshold, sending a prompt to the teacher that "students' attention has decreased, interaction is recommended"; When a prolonged period of no response is detected, a prompt is sent to the teacher suggesting that they adjust their questioning style or pace. When a teacher speaks too fast, a notification will be sent to the teacher saying "Speaking too fast, it is recommended to slow down"; When active student participation is detected, a positive incentive signal is sent to the classroom equipment.

6. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The classroom brain uses a domestically produced edge computing chip with a built-in tri-core NPU, providing 20 TOPS of AI computing power and supporting real-time decoding of 32 channels of 1080P video streams, enabling localized processing of classroom data and ensuring that sensitive teaching data does not leave the classroom.

7. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The campus superbrain adopts a domestically developed information technology architecture, including the Kunpeng processor and the Ascend AI accelerator card. It has a built-in controllable data set hub, a campus business application hub, and a one-stop intelligent agent hub, and supports local fine-tuning and training of education-specific models.

8. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, It also includes a fully visualized teaching and research robot, which supports administrators or teaching and research groups to remotely listen to classes, observe teaching methods and student participation, and provide multi-dimensional AI analysis reports.

9. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, It also includes an automatic audio and video segmentation robot, which is used to automatically segment recorded videos according to classroom knowledge points and generate short videos of knowledge points for students to review and teachers to use for teaching and research.

10. The proactive AI-powered intelligent teaching feedback system based on classroom behavior data analysis according to claim 1, characterized in that, The class brain and the campus super brain adopt a closed-loop data flow mechanism: the class brain collects and preprocesses data, uploads it to the campus super brain for model training, the campus super brain distributes the updated model, the class brain applies the model to generate feedback and the feedback effect data is sent back to the campus super brain, forming a continuous optimization closed loop.

Images