64 results about "Activity classification" patented technology

The present invention relates to an intelligent, context-aware battery management system embedded within a mobile device that dynamically allocates power resources based on real-time user behavior, system state, and environmental context. It incorporates a smart recognition engine that analyzes sensor-derived telemetry data to compute behavioral deviation scores, enabling the system to anticipate abnormal or emergency-prone conditions. A continuous activity classification module contextualizes user motion and geolocation to inform power policy decisions. Upon detecting significant behavioral anomalies or critically low battery conditions, an emergency mode subsystem is triggered, restricting device operations to essential functionalities while preserving energy for critical communication and navigation tasks. The system also establishes a secure, lightweight emergency communication tunnel for relaying essential metadata, including GPS and behavioral indicators, to predefined response servers.

A system and method for person identification from video data by disentangling biometric identity features from non-biometric appearance and activity features are disclosed. The system processes RGB video sequences depicting individuals performing various activities to extract spatio-temporal features. These features are separated into distinct biometric identity representations and non-biometric features related to appearance and performed activities. To achieve this separation and minimize appearance bias, the system utilizes an auxiliary supervisory model. At least two implementations of this supervisory model are disclosed: one using semantic supervision via structured embeddings processed through a vision-language model, and another employing silhouette-based feature distillation from a silhouette-trained neural network. Joint training for biometric identification and activity classification ensures accurate identification of individuals independently of facial visibility, clothing differences, or activity variations.

Techniques and apparatuses are described that perform audioplethysmography and motion-sensing data fusion. Fusing motion-sensing data with audioplethysmography expands the situations in which audioplethysmography can operate. In one aspect, audioplethysmography and motion-sensing data fusion can be used for motion-artifact filtering. With motion-artifact filtering, noise caused by motion of a user can be attenuated to improve sensitivity and accuracy for audioplethysmography. This performance improvement expands the ability of audioplethysmography to support use cases during situations in which the user engages in an activity. In another aspect, audioplethysmography and motion-sensing data fusion can expand a functionality of a hearable to include activity detection and / or activity classification. While activity detection and / or activity classification can provide additional contextual information for other use cases associated with audioplethysmography, either can also be used to control an operation of the hearable and / or a computing device.

Various embodiments provide a system, method, and device for applying a DNS activity classification framework. The method incudes (i) collecting DNS-related activity, (ii) determining whether the DNS-related activity is associated with a DNS tunneling campaign or tool, and (iii) performing an active measure in response to detecting DNS traffic associated with a tunneling domain.

Techniques and apparatuses are described that perform audioplethysmography and motion-sensing data fusion. Fusing motion-sensing data with audioplethysmography expands the situations in which audioplethysmography can operate. In one aspect, audioplethysmography and motion-sensing data fusion (202) can be used for motion-artifact filtering (206). With motion-artifact filtering (206), noise caused by motion of a user (106) can be attenuated to improve sensitivity and accuracy for audioplethysmography. This performance improvement expands the ability of audioplethysmography to support use cases during situations in which the user (106) engages in an activity. In another aspect, audioplethysmography and motion-sensing data fusion (202) can expand a functionality of a hearable (102) to include activity detection (208) and / or activity classification (210). While activity detection (208) and / or activity classification (210) can provide additional contextual information for other use cases associated with audioplethysmography, either can also be used to control an operation of the hearable (102) and / or a computing device.

Methods, apparatus, and processor-readable storage media for determining security risks related to local administrator rights (LAR) activity are provided herein. An example computer-implemented method includes obtaining data pertaining to one or more activities performed by at least one user acting in connection with at least one granted set of LAR; classifying the one or more activities into one or more security risk-based categories by processing at least a portion of the obtained data; determining one or more security-related recommendations based at least in part on the classifying of the one or more activities into the one or more security risk-based categories; and performing at least one automated action based at least in part on at least a portion of the one or more security-related recommendations.

A computer-enabled system and method that collects and processes various data from mobile devices possessed by on-demand cargo delivery transportation drivers, along with other possible external data sources, and subsequently recognizes, classifies, stores, and delivers cargo delivery-related activity information of individual or multiple drivers to various consumers of the system, while also iteratively improving the system and method's cargo delivery activity classification accuracy over time.

The invention discloses a Crohn disease focus automatic segmentation and activity evaluation system based on deep learning, which belongs to the field of medical artificial intelligence and comprises a data preprocessing unit, a focus automatic segmentation unit, a radiomics feature extraction unit, a feature screening and dimension reduction unit and an activity classification unit. According to the method, an nnU-Net deep learning segmentation model is combined with image omics feature extraction, multi-stage feature screening and machine learning classification technologies, so that full-process automation from CTE image preprocessing, focus automatic segmentation, feature extraction and screening to activity classification is realized. The system can efficiently and accurately segment the focus of Crohn's disease, automatically assesses the disease activity based on the screened key radiomics characteristics, significantly improves the consistency, objectivity and efficiency of diagnosis, and is suitable for clinical auxiliary diagnosis and scientific research analysis.

The invention provides a radar-based full-time in-bed state monitoring method, system and product, and the method comprises the steps: carrying out the signal preprocessing of a collected radar echo signal, and obtaining target existence information and target physiological information; performing feature extraction on the target existence information and the target physiological information, and adaptively dividing a daytime period and a night period based on a feature extraction result; performing multi-mode sleep state staging based on the target existence information and the target physiological information in the night time period; performing daytime activity state classification based on the target existence information and the target physiological information in the daytime period; and generating a full-time in-bed state monitoring report based on the sleep state staging result and the daytime activity state classification result. According to the method provided by the invention, both long-term bedridden people and non-bedridden people can be considered, the day and night time periods are automatically switched, and the function that single radar equipment synchronously supports night sleep staging and daytime activity classification is realized.

The invention provides a cold start method of a brain-computer interface, which relates to the technical field of brain-computer interfaces, and comprises the following steps: collecting an initial signal of a target user by using brain-computer interface equipment; the initial signal comprises electroencephalogram data and magnetoencephalogram data; preprocessing the initial signal to obtain a preprocessed signal; performing feature extraction on the preprocessed signal to obtain a signal feature; inputting the signal features into a trained electroencephalogram activity classification model to obtain a classification result; and generating a feedback result according to the classification result, and sending the feedback result to the target user. The cold start method of the brain-computer interface effectively solves the problems that a system based on transfer learning is poor in adaptive capacity and slow in response in the cold start stage of a new user. According to the method, individuation, real-time feedback and online learning are emphasized, so that a more effective and flexible user adaptation scheme is provided, the interactive experience of the user is smoother, and the overall performance of the system and the user satisfaction degree are improved.

A system and method to obtain activity data for an activity of a user, the activity data corresponding to multiple parameters. A set of threshold criteria that corresponds to a multi-parameter activity zone is retrieved, where the set of threshold criteria includes an effort threshold criteria related to a heart rate of the user. An effort of the user is determined based on the activity data. A resistance experienced by the user is determined, where the resistance includes a measure representing at least one of stride rate, step rate, steps, or cadence. One or more environmental conditions associated with the user are determined. Based on the set of threshold criteria corresponding to the multi-parameter activity zone, an instance of the activity is classified as an activity type based on the resistance experienced by the user, the effort of the user, and the one or more environmental conditions associated with the user.

The invention discloses a human body activity identification method and system capable of effectively capturing variable time-space relations in sensors and between the sensors. The method comprises the following steps: 1) carrying out cross-variable fusion on local cross-channel fusion features; 2) performing global time aggregation on the local cross-variable fusion features to obtain sensor feature data; 3) integrating different sensor feature data by using a self-attention mechanism to generate human body activity features; and 4) inputting the human body activity features into a full-connection linear classifier to obtain human body activity categories. The system comprises N wearable motion sensors, a data conversion module, a local time feature extraction module, a cross-channel fusion module, a cross-variable fusion module, a global time aggregation module, a human body activity feature extraction module and a human body activity classification module. By capturing the space-time relationship between the interior of the sensor and the sensors, the frame can provide more systematic information, so that the model is more accurate during analysis and prediction.

An activity classifier system and method that classifies human activities using 2D skeleton data. The system includes a skeleton preprocessor that transforms the 2D skeleton data into transformed skeleton data, the transformed skeleton data comprising scaled, relative joint positions and relative joint velocities. The system also includes a gesture classifier comprising a first recurrent neural network that receives the transformed skeleton data, and is trained to identify the most probable of a plurality of gestures. The system also has an action classifier comprising a second recurrent neural network that receives information from the first recurrent neural networks and is trained to identify the most probable of a plurality of actions.

Systems and methods include acquisition of a first image of a first activity record, determination of first text based on the first image, generation of a first embedding based on the first text, generation of a second embedding based on the first embedding using a first model, where a number of dimensions of the second embedding is less than a number of dimensions of the first embedding, determination of a first cluster based on the second embedding using a second trained model, the second trained model trained using unsupervised learning, and determination of a first activity type associated with the first activity record based on the first cluster, the second embedding and historical activity data associating the first cluster with a plurality of activity types and each of the plurality of activity types with a respective embedding metric.

Medical systems, methods, computer-readable media, and program products with physiological sensors are provided. In an example method, a computer system receives acceleration data, activity classification data, sleep data, and one or more vital sign metrics indicative of physical activity of a subject from a wearable medical sensor. The system determines a physical activity score, a sleep score, and a vital sign score based on acceleration data, activity classification data, sleep data, and one or more vital sign metrics. The system determines a quality of life score for the subject based on the physical activity score, the sleep score, and the vital sign score; and the system causes the quality of life score to be presented to the subject using the display screen.

This application discloses a method, apparatus, and electronic device for determining user activity. The method includes: acquiring first feature information reflecting the activity of any user; clustering a first user group according to the first feature information and a preset clustering algorithm to obtain at least one user cluster, wherein each user cluster corresponds to a certain level of activity; determining multiple activity classification criteria according to the at least one user cluster and a preset classification model, wherein each activity classification criterion involves constraints on the first feature information; and determining the activity of a target user according to a target activity classification criterion among the multiple activity classification criteria.

Implementations disclosed describe methods, devices, and systems to perform out-of-distribution and recognition of activities using an inertial measurement unit (IMU) sensor. A method may include receiving motion data by a device from a motion sensor. The method further includes generating image data comprising one or more images based on the motion data. The method further includes determining that a first portion of the image data corresponds to activities outside a classification distribution. The method further includes filtering the image data by removing the first portion from the image and generating filtered image data. The method further includes determining an activity classification, within the classification distribution, based on the filtered image data. The method further includes modifying an operating parameter of the device based on the activity classification.

Methods, systems, and devices for activity classification are described. A system may receive physiological data associated with a user via a wearable device, where the physiological data includes at least motion data. The system may identify an activity segment during which the user is engaged in a physical activity based on the motion data, where the activity segment is associated with activity segment data including at least the physiological data collected during the activity segment. The system may generate activity classification data associated with the activity segment based on the activity segment data, the activity classification data including a set of classified activity types and corresponding confidence values. The system may then cause a graphical user interface (GUI) of a user device to display the activity segment data and at least one classified activity type of the set of classified activity types.

The present invention provides a multi-action torque assistance control method for a knee exoskeleton based on activity classification and phase detection, including a two-stage artificial neural network and three human body dynamics models; the two-stage artificial neural network, the first stage consists of a network for activity classification, and the second stage consists of three independent gait phase estimation regression networks. The three human body dynamics models include a single-support human body model, a double-support human body model, and a swing human body model divided according to the human gait phase characteristics. The model outputs the true biological joint torque at the current moment, and the controller proportionally outputs the required auxiliary torque curve of the exoskeleton. The present invention can accurately classify four activity types of standing, walking, going up stairs, and going down stairs, accurately estimate the gait phase percentage under non-rhythmic gait conditions, and generate corresponding auxiliary torque curves based on the current gait category and gait phase.

One or more computing devices, systems, and / or methods are provided. In an example, a method comprises receiving, by a device, incoming motion data from a motion sensor, generating, by the device, an incoming embedding vector based on the incoming motion data, generating, by the device, a predicted embedding vector based on the incoming embedding vector, assigning, by the device, an activity classification based on the predicted embedding vector, and modifying an operating parameter of the device based on the activity classification.

The embodiment of the invention provides a visual and auditory processing and brain activity classification method and system based on EEG space-time-frequency characteristics. The method comprises the steps that multi-band electroencephalogram data are obtained; wherein the multi-frequency-band electroencephalogram data represents brain activity data collected under different frequency bands; performing visual and auditory preprocessing on the multi-band electroencephalogram data to obtain a preprocessed multi-band electroencephalogram signal; performing clustering analysis on the basis of the multi-frequency-band electroencephalogram signals to obtain micro-states of various stimuli under different frequency bands; calculating a micro-state attribute of the micro-state, and performing relevance calculation based on the micro-state attribute and a predetermined multi-band brain network to obtain a space-time frequency relevance vector; and through a predetermined visual and auditory processing brain activity identification model, carrying out brain activity identification on the space-time-frequency association vector to obtain a brain activity classification result. According to the scheme, the brain activity classification accuracy can be improved.

An activity classifier system and method that classifies human activities using 2D skeleton data. The system includes a skeleton preprocessor that transforms the 2D skeleton data into transformed skeleton data, the transformed skeleton data comprising scaled, relative joint positions and relative joint velocities. The system also includes a gesture classifier comprising a first recurrent neural network that receives the transformed skeleton data, and is trained to identify the most probable of a plurality of gestures. The system also has an action classifier comprising a second recurrent neural network that receives information from the first recurrent neural networks and is trained to identify the most probable of a plurality of actions.

Systems and methods for scraping detection include a device which receives a plurality of requests from a client to extract data from a resource. The device may classify activity of the client as activity of an autonomous program based at least on a number of the plurality of requests, and one of i) one or more content types of the requests, or ii) a frequency in which the requests are received. The device may block, responsive to classification of the activity, a subsequent request from the client to extract data from the resource.

Systems and methods are provided to classify activities into clusters. The systems and methods access data representing activity on a network site and generate, based on the data, a plurality of feature sets representing periods of activeness on the network side. The systems and methods form a subset of the plurality of feature sets by reducing dimensionality of the plurality of feature sets. The systems and methods generate a plurality of clusters of the subset of the plurality of feature sets, each cluster being associated with a label representing a different type of activeness on the network site, and generate a database query set based on the plurality of clusters and the plurality of feature sets to classify one or more activities on the network site into one of the plurality of clusters.

A method of determining an event participated in by an athlete includes receiving, at a computing device, a plurality of motion determinations generated by a monitor from motion data captured from the athlete's motions during a monitoring window. The motion determinations include actions performed by the athlete, performance metrics of the athlete, or both. The method also includes classifying, by use of a machine learning model stored on the computing device, which event among a plurality of predetermined events the athlete participated in during the monitoring window based at least in part on the plurality of motion determinations.

The application provides a portable full-automatic fluorescence spectrum sperm activity detection method and system, and belongs to the technical field of sperm activity detection. The method comprises the following steps: placing a semen sample to be detected into a detection kit to which a sperm activity fluorescence biological dyeing agent is added; the sperm activity fluorescence biological dyeing agent comprises a CM-Dil biological derivative activation dyeing solution; fluorescence spectrum information of the semen sample to be detected is acquired through a far-field optical fluorescence microscope and a microscopic optical information acquisition assembly; the fluorescence spectrum information is subjected to data analysis through a spectrum analysis system, so as to obtain survival rate information, development state information and activity classification information of sperm in the semen sample. The application realizes that a person to be detected can complete sperm activity detection by himself without going to a hospital; and the detection result contains the survival rate information, the development state information and the activity classification information of sperm in the semen sample, so that whether the person to be detected has the ability to reproduce can be accurately determined.

The application relates to the technical fields of computer vision, sensor signal processing and artificial intelligence, and specifically provides a low-power millimeter wave radar human behavior recognition method and system, which comprises the following steps: preprocessing point cloud data collected by a millimeter wave radar to obtain a structured feature sequence; extracting multi-level manual features from the structured feature sequence, and performing feature selection and normalization to generate an optimized feature vector; inputting the optimized feature vector into a heterogeneous integrated classifier, combining a dynamic pseudo-label strategy to perform semi-supervised training on the heterogeneous integrated classifier, and obtaining an enhanced recognition model; and using the enhanced recognition model to perform human activity classification and recognition on millimeter wave radar data to be recognized. The application significantly improves the recognition accuracy and robustness of behavior type recognition from millimeter wave radar data.

Disclosed are systems, methods, and apparatus for unsupervised training of an activity classifier for movement measured at a target location of a user using a trained source activity classifier configured to determine an activity based on motion measured at a source location of a user. According to exemplary embodiments, the activity classifier for the target location can be trained using source embeddings generated by a trained classifier for the source location. For example, the source embeddings and the motion measured at the target location may then be provided as training inputs to the target activity classifier for the target location of a user, so that the target activity classifier can reconstruct the source embedding for the target motion without the use of any labeled training data and / or without having to learn the transfer function of the existing trained machine learning model.

A method of determining events participated by an athlete includes receiving, at a computing device, a plurality of motion determinations generated by a monitor from motion data captured from a motion of the athlete during a monitoring window. Athletic determination includes actions performed by the athlete, performance metrics of the athlete, or both. The method further includes classifying which of the plurality of predetermined events the athlete participates during the monitoring window, by using a machine learning model stored on the computing device, based at least in part on the plurality of motion determinations.

A method and system for activity classification. A pressure sensor receives input data resulting from physical activity of a subject performing an activity. The input data includes pressure data from at least one pressure sensor, and may include other data acquired through other types of sensors. A deep learning neural network is applied to the input data for identifying the activity. The neural network is trained with reference to training data from a training database. The training data may include empirical data from a database of previous data of corresponding activities, synthesized data prepared from the empirical data or simulated data. The training data may include data from physical activity of the subject being monitored by the system. Different aspects of the neural network may be trained with reference to the training data, and some aspects may be locked or opened depending on the application and the circumstances.