462 results about "Head posture" patented technology

In various examples, head pose prediction for automotive occupant sensing systems and applications is presented. The systems and methods described herein provide for a machine learning model trained using a dataset that comprises ground truth head pose data computed using a registered head model of a training subject. While operating a vehicle, one or more cameras and a depth sensor capture synchronized images of the training subject. To compute a ground truth 3D head pose, angular deviations between a 3D point cloud and the registered head model may be computed to obtain a 3D ground truth head pose measurement. Using an extrinsic calibration transform, the head pose measurement may be mapped into the sensor coordinate frame. Training samples may be produced for training the machine learning model that comprise an optical image frame and the head pose measurement transposed into the frame of reference for that optical image frame.

The invention provides a driving behavior intelligent identification method and system based on video analysis, and the method comprises the steps: obtaining a driver face video stream and a road environment video stream collected by a vehicle-mounted camera, and reading the driving information recorded by a whole vehicle communication network; recognizing an eyelid closing state, a sight line direction and a head posture in the driver face video stream based on a posture recognition model, and performing fatigue distraction analysis to obtain driver state information; performing motion trail analysis on the road environment video stream and the driving information, and performing driving risk assessment in combination with the driver state information to obtain driving assessment information; and performing early warning construction according to the driving evaluation information, generating early warning prompt information, and synchronously writing the early warning prompt information, the driving evaluation information and the driver state information into a safety data protection memory. The fatigue and distraction states of the driver can be recognized more accurately, and the accuracy of state judgment is improved.

The invention relates to the technical field of computers, and discloses an interaction control method of intelligent glasses. The method comprises the following steps: synchronously acquiring multi-modal data such as eye movement, voice, gestures and head postures and environment and application context information; carrying out independent time sequence feature coding on each modal data; generating a modulation vector in combination with the context, and outputting probability distribution of user intentions through a cross-modal attention fusion network; and a unique execution instruction is determined through an instruction arbitration module based on rules and a state machine. The system comprises corresponding function modules. According to the method, the accuracy, robustness and naturalness of interaction are improved through multi-modal synchronous fusion and a context self-adaption mechanism.

The invention provides a target distance estimation method and system based on optical characteristics of a holder, and the method comprises the steps: carrying out the automatic target detection and tracking shooting of a to-be-detected target at a current visual angle, and obtaining a target video of the to-be-detected target; meanwhile, the cradle head posture and the camera focal length are collected in real time; performing target detection according to the target video to obtain key feature points of the target; performing distortion correction on the key feature points through image coordinate mapping to form a unified image coordinate system; combining the image coordinate system with the attitude information of the holder, and mapping the image coordinate system to a three-dimensional virtual space to construct a virtual coordinate system; and calculating the target distance of the to-be-measured target according to the geometric position of the to-be-measured target in the virtual coordinate system. According to the invention, the target distance can be accurately estimated in a complex scene and at different visual angles based on the holder system with a dynamic attitude adjustment function, and the method has the technical effects of real-time performance and robustness.

The invention provides a head posture estimation method and system based on a neural implicit field, and belongs to the technical field of image processing. According to the method, discretization processing is carried out on a head posture label, and a Gaussian function and a Gumbel function are used for joint modeling to generate an asymmetrically distributed label. Sampling points are arranged at fixed intervals in a three-dimensional space, and global features of an input RGB image are extracted through a RepVang network. And establishing a mapping relationship between image features and spatial sampling points by means of a neural implicit field, mapping face information to a point cloud space, and performing binary classification on each sampling point by using a multi-layer perceptron to generate human face point cloud data. Geometric features of the point cloud data are extracted and fused with global features of the RGB image, and the head posture is predicted through combination of a classification module and a regression module. By adopting the method and the system, the precision and the robustness of head posture estimation can be effectively improved, and a reliable solution is provided for practical application.

Examples of wearable systems and methods can use multiple inputs (e.g., gesture, head pose, eye gaze, voice, totem, and / or environmental factors (e.g., location)) to determine a command that should be executed and objects in the three-dimensional (3D) environment that should be operated on. The wearable system can detect when different inputs converge together, such as when a user seeks to select a virtual object using multiple inputs such as eye gaze, head pose, hand gesture, and totem input. Upon detecting an input convergence, the wearable system can perform a transmodal filtering scheme that leverages the converged inputs to assist in properly interpreting what command the user is providing or what object the user is targeting.

The invention discloses a distance measurement system based on artificial intelligence, and the system can achieve the automatic grabbing of size parameters of glasses. According to the system, a camera is adopted to capture a glasses image, a gyroscope and an artificial intelligence algorithm are adopted to recognize the spatial position of glasses, spatial coordinates of feature points are obtained, the pixel scale # imgabs0 # of the image is obtained subsequently based on the physical geometrical relationship of the glasses and the real pupil distance of a user, and finally the size parameters of the glasses are obtained. The glasses parameters comprise a glasses width, a pupil distance, a front inclination angle, a surface bending angle and a glasses-eye distance; the camera is an apple panel built-in camera; the gyroscope can obtain the pitch angle # imgabs1 #, the deflection angle # imgabs2 # and the inclination angle # imgabs3 #, so that the head posture of a glasses wearer is obtained, and the size measurement precision of the glasses is greatly improved. According to the system, automatic measurement of glasses parameters is realized, the working intensity of manual glasses matching is greatly reduced, and digital upgrading of high-end customized glasses is realized.

The invention relates to an automatic emergency braking threshold value adjusting method and system based on in-cabin multi-mode information, and relates to the technical field of auxiliary driving, the method comprises the steps that visual information and audio information of a driver and at least one passenger in a cabin are obtained, the visual information comprises facial expressions and head postures, and the audio information comprises audio information of the driver and the at least one passenger; the audio information comprises voice features and non-voice sounds; respectively calculating a driver state coefficient and a passenger state coefficient based on the visual information and the audio information; fusing the driver state coefficient and the passenger state coefficient to obtain an in-cabin safety situation coefficient; and according to the in-cabin safety situation coefficient, a triggering threshold value of the automatic emergency braking system is adjusted in a self-adaptive mode. According to the method, multi-modal information such as vision and audio in the cabin is fused, and state analysis of all passengers is combined, so that intelligent self-adaptive adjustment of the AEB braking threshold value can be realized, and the sensing accuracy and decision reliability of an emergency braking system in a complex scene are improved.

The invention relates to the technical field of intelligent glasses, and discloses an intelligent glasses wearing identification method based on multi-modal perception, and the method comprises the steps: collecting the capacitance change rate, the facial contour matching degree, the body surface temperature, the vital signs, the head attitude angle and other multi-source feature data in real time, generating five groups of feature data, and carrying out the standardization processing and weight fusion, forming a comprehensive feature vector; calculating a preliminary confidence score by combining a dynamic weight adjustment mechanism and a rule engine, and correcting a deviation caused by head movement; judging a wearing state according to the corrected confidence score, outputting a label, and realizing function control in linkage with a gesture interaction module; the abnormal state triggers local or cloud alarm; the sampling frequency of the sensor is dynamically adjusted according to the wearing state, the frequency is reduced to 1Hz when the sensor is not worn, the frequency is improved to 10Hz when the sensor is worn so as to enhance the response precision, and the decision process is only triggered when the feature is suddenly changed in combination with an event-driven strategy. According to the invention, the precision and the reliability can be improved, and low-power-consumption design is realized.

The invention discloses a data processing and interaction control system based on dynamic eye position tracking, and particularly relates to the technical field of man-machine interaction and eye movement tracking, which comprises an image acquisition module, a data processing module, an interaction control module and an equipment interface module. The image acquisition module controls a near-infrared camera to capture an eye video; the data processing module executes dynamic calibration, triggers calibration points based on a historical gaze point distribution sparse region, updates a sight line mapping model online by adopting a recursive least square method, and recognizes an interaction intention through a state machine in combination with a gaze point sequence and an eye movement event mode; the interaction control module carries out logic binding on the eye movement intention and signals such as head postures, gestures and voice, and a composite control instruction is generated through complementary fusion, concurrent binding and conflict arbitration; and the device interface module standardizes the instruction and then outputs the instruction to the controlled device. The system realizes non-intrusive dynamic calibration, effectively improves tracking precision and interaction robustness, and is suitable for a multi-modal fusion control scene.

The invention provides a classroom student concentration degree monitoring system and method based on computer vision. The system comprises a vision collection module, an edge calculation unit and a cloud analysis platform. A multi-dimensional sensing mode combining head posture analysis, eye state tracking, facial expression recognition and body movement detection is adopted to evaluate the concentration degree, through a dynamic feature fusion mechanism, the system can intelligently adjust the weight ratio of different modal features, the limitation of single-dimensional analysis in a traditional method is effectively solved, and the accuracy of the system is improved. The system adopts edge calculation and cloud analysis, ensures high efficiency of data processing, ensures reliability of complex algorithm operation, and has important value for improving classroom teaching quality and learning efficiency. According to the invention, through multi-modal visual analysis, behavior characteristics and physiological reactions of students in class are monitored in real time, the concentration level of the students is comprehensively evaluated, and objective teaching feedback and improvement basis are provided for teachers.

The invention relates to the technical field of automobile safety detection, and discloses a fatigue driving real-time detection method and system based on a monitoring camera. The method comprises the following steps: acquiring a multi-source sensing data stream in a face area of a driver, and performing time sequence alignment processing to generate a synchronized physiological signal set; multi-modal feature extraction is carried out on the face muscle activity intensity image to obtain an eye movement track feature vector, a head posture change sequence and a face muscle activity intensity map; respectively generating an eye movement fatigue index and a head behavior anomaly score, and obtaining a comprehensive fatigue grade index in combination with a weighted decision algorithm; generating a graded early warning trigger strategy according to the comprehensive fatigue grade index and the real-time driving environment parameter, and calling a corresponding alarm instruction to generate a multi-stage warning signal output stream; and monitoring the change trend of the physiological state of the driver in real time, carrying out dynamic priority adjustment on a warning signal output flow, updating an alarm instruction set, and completing fatigue driving intervention closed-loop control.

The invention belongs to the field of intelligent driving interaction, discloses an HUD interaction method and system combining DMS and ADAS functions, and constructs a multi-mode state sensing system by obtaining a real-time image of a driver and steering wheel grip data. According to the method, eye, pupil and head posture recognition based on deep learning is complementary with a grip distribution matrix, so that fatigue behaviors, distraction behaviors and abnormal holding behaviors can be detected earlier and more accurately, and the precision and robustness of driver state recognition are effectively improved. According to the method, after the fatigue degree of the driver is analyzed, the cognitive load of the driver is dynamically evaluated in a finer-grained mode by constructing the composite indexes based on the eyelid opening degree, the sight line deviation, the grip variation coefficient and the respiratory rhythm. The evaluation mode can reflect the attention, mental load and information processing capability of the driver in real time, so that the system can generate a prompt strategy which better conforms to human factors according to the current state of the driver.

The invention discloses a classroom interaction optimization method and system driven by a teaching posture thermodynamic diagram, and relates to the technical field of education informatization. Comprising the following steps: acquiring image data acquired by a classroom multi-angle camera; extracting a teacher skeleton point coordinate sequence and a student head posture vector; determining a teacher movement track and a teacher gesture unfolding angle parameter; mapping to a classroom plane coordinate system in a ray projection mode to generate a student watching focus coordinate set; performing correlation analysis on the teacher movement track and the student watching focus coordinate set to obtain a time-space correlation parameter; inputting a time sequence prediction model to generate an attention change rate, and taking an attention change rate time point exceeding a preset change threshold as attention attenuation inflection point time; and generating teaching movement line optimization suggestions and outputting visual prompt information. According to the invention, teaching movement line optimization based on teacher-student interaction characteristics can be realized, and classroom interaction efficiency is improved.

The present disclosure relates to a multi-mode vehicle window control method and device. The method combines voice recognition and computer vision technologies, detects a user's voice wake-up word using a vehicle-mounted microphone, turns on the vehicle-mounted camera to detect the user's face and head, analyzes the user's gaze, face pose, and head posture to estimate the user's attention region, and aligns the attention region with a vehicle body coordinate system. Intelligent control of a vehicle window is realized according to specific coordinates of an attention area on the vehicle window. Even if the user's face or eyes are not detected, the vehicle window may be operated accurately by determining the user's intention through the face or head pose.

The invention provides a head-eye posture fused driver state grading intervention method and system, and the method comprises the steps: dividing a fixation region which comprises a safety region, an alarm region and a dangerous region based on an in-vehicle three-dimensional coordinate system; calculating an initial line-of-sight vector according to the driver face key points, obtaining a final line-of-sight vector by combining line-of-sight deviation angle correction, and obtaining a fixation point coordinate based on the eye point coordinate and the final line-of-sight vector; based on fixation point coordinates and fixation area matching, the driver fixation area category is judged; and acquiring the current vehicle speed and the fixation duration in the fixation area, inputting the current vehicle speed and the fixation duration into a pre-trained decision model to obtain a graded intervention strategy, and executing an intervention action according to the strategy. The accuracy of the fixation point position is ensured through fusion of the head posture and the sight line deviation; a grading intervention strategy is obtained by combining the fixation point position, the fixation duration and the vehicle speed, interference to normal driving is reduced, effective early warning can be achieved in the risk accumulation stage, and the method has adaptability to complex vehicle use scenes.

The invention relates to the technical field of video communication, in particular to an intelligent AI glasses video call auxiliary system which comprises a visual adjustment module, a voice interaction module, a communication scheduling module, a posture recognition module and a feedback prompt module. According to the invention, by acquiring the eye movement track and the pupil diameter change, adjusting the image brightness and the color contrast in combination with the ambient light, relieving visual fatigue, enhancing the call comfort, analyzing the voice frequency band and matching the keywords, the accurate recognition of the operation intention is realized, the call response efficiency is improved, and the mistaken touch and delay are avoided; the call priority is judged according to the action state, the acceleration trend and the contact person weight, the task access accuracy is improved, the user intention is judged according to the head posture and the sight line trend, misoperation and repeated confirmation are avoided, interaction convenience is enhanced, meanwhile, the audio intensity and image prompt are regulated and controlled according to the task type and the prompt level, the multi-environment prompt requirement is met, and the user experience is improved. Information omission is reduced, and call experience is optimized.

The invention provides intelligent glasses, a shooting control method, electronic equipment and a storage medium, and the intelligent glasses are characterized in that the lower end of a miniature stable assembly is arranged on a glasses frame, a camera module is installed at the front end of the glasses frame through the upper end of the miniature stable assembly, and the miniature stable assembly is used for adjusting the shooting angle of the camera module. The micro stabilizing assembly has at least two rotational degrees of freedom which are perpendicular to each other; the posture sensor is arranged on the glasses frame and used for detecting head posture changes of a wearer in real time and outputting posture data; and the control processing module is electrically connected with the attitude sensor and the micro stable assembly, and is used for receiving the attitude data, calculating an angle correction value based on the attitude data, and driving the micro stable assembly to dynamically adjust the shooting angle of the camera module based on the angle correction value. Therefore, the shooting angle is always consistent with the actual sight direction of the wearer. Viewing offset caused by wearing slippage and posture change is effectively eliminated, and picture instability is remarkably improved.

The invention discloses a multifunctional automatic comprehensive optometry combined operation table and a use method thereof. The operating platform comprises a mechanical main body structure, a posture sensing system and a central processing and control unit. The mechanical main body structure comprises an optometry chair driven by a rotating disc, a lifting mechanism and an electric rotating shaft, and the posture sensing system is composed of a sitting and leaning posture recognition mechanism used for recognizing body postures and a head posture recognition mechanism used for capturing head movement. The central processing and control unit judges the posture of the optometry person according to the sensing data, and if the posture does not meet the preset standard, the seat height, the backrest angle and the horizontal position are adjusted through voice prompt and an automatic driving mechanism; when the posture reaches the standard, the optometry equipment is automatically controlled to complete operation, automatic and accurate correction of the posture before optometry and intelligentization of the optometry process are achieved, and the accuracy and efficiency of optometry results are effectively improved.

The invention provides a voice processing method and device and XR equipment, and relates to the technical field of voice processing, and the method is applied to augmented reality XR equipment, and comprises the steps: obtaining a current scene image and the current head posture data of a user; determining a target watching object according to the current scene image and the current head posture data; acquiring real-time audio data and real-time video data of the target watching object; and processing the real-time audio data and the real-time video data through a voice separation model, and determining a target audio signal of the target watching object. According to the invention, the voice signal of the target speaking object can be accurately recognized on the premise of not increasing the extra hardware cost of the XR equipment and not depending on the prior information of the target speaking object.

The invention relates to the technical field of AR interaction control, in particular to an AR glasses cursor control method and system based on head posture induction, and the method comprises the steps of data acquisition, mapping establishment, displacement accumulation and cursor control. Through the AR glasses equipment body, the natural rotation posture of the head of the user is collected and used as an input instruction, and a reverse mapping normal form is constructed, so that the physical direction of cursor movement is opposite to the physical direction of user head rotation, and a non-intuitive chasing relation between the cursor and a target is avoided; intuitive mapping between natural rotation of the head of the user and cursor control of the AR glasses is achieved, accurate control over the cursor can be achieved without additional auxiliary actions or a controller, and convenience and privacy are high.

The invention relates to the field of artificial intelligence, in particular to a depression tendency detection method and device based on a dynamic three-dimensional face. According to the method, dynamic three-dimensional face sequence data is adopted as a basis, dual-path feature extraction of a depth mode and a texture mode and head posture frequency domain analysis are fused, and multi-mode feature adaptive fusion is realized through an attention mechanism. The technical scheme is the first time in the field of depression tendency detection, and can effectively overcome the problems of insufficient utilization of depth information, incomplete dynamic mode capture, lack of head signal analysis and the like in the prior art. An objective, dynamic and efficient quantification method is provided in preliminary screening of depression tendency, fine behavior characteristic changes can be captured, and the method is suitable for rapid screening of large-scale crowds.

The embodiment of the invention provides a speaking face video generation method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring a target audio and a target face image, extracting audio coding features of the target audio through a target audio encoder, predicting an object head posture sequence based on the audio coding features through a target head posture predictor, and performing feature fusion on the audio coding features and the object head posture sequence to obtain a target head posture image; carrying out forward diffusion processing on the target face image to obtain a noise action sequence, carrying out reverse diffusion processing according to the target fusion feature and the noise action sequence through a target action generation model to obtain a target action sequence, and extracting a target appearance hidden variable of the target face image through a target image encoder; the target speaking face video is generated through the target decoder based on the target action sequence and the target appearance hidden variable, and the quality of speaking face video generation can be improved.

The invention relates to the related technical field of student behavior monitoring, in particular to a classroom student behavior monitoring system, which comprises the following functional modules: an identity label generation module; an image acquisition module; a data storage module; a behavior analysis module; a multi-modal data integration module; and an interaction feedback module. According to the classroom student behavior monitoring system, the identity label generation module, the image acquisition module, the data storage module, the behavior analysis module, the multi-modal data integration module and the interaction feedback module work cooperatively; algorithm design of a head posture analysis unit, a hand action recognition unit, an electronic equipment use detection unit and a sitting posture evaluation unit is further optimized, and a student behavior track diagram and an abnormal condition report are provided as dynamic monitoring results to an external user by means of an interaction feedback module. The defects of the prior art in the aspects of recognition precision, real-time performance, multi-dimensional data analysis and privacy protection are overcome.

The invention discloses a driver fatigue detection optimization method based on a Transform-YOLO (Young Line Offset). Comprising the following steps of data acquisition, data preprocessing, target detection, key region feature extraction, multi-modal feature fusion and fatigue state classification. The method comprises the following steps: firstly, detecting a face and key areas in a video stream of a driver through a YOLOv11 model, wherein the face and key areas comprise postures of eyes, a mouth and a head; the method comprises the following steps: firstly, extracting a target detection result, then extracting spatio-temporal characteristics by utilizing a key region attention (KRA) module, and splicing the target detection result and the spatio-temporal characteristics in combination with a multi-modal characteristic fusion technology; and finally, realizing fatigue state judgment through a classifier. According to the method, the problems of calculation redundancy and insufficient real-time performance of the traditional Transform in feature modeling of the driver face key areas (eyes and mouths) are solved.

The invention discloses a control method and device of intelligent glasses, the intelligent glasses and a storage medium, the method is applied to the intelligent glasses, the intelligent glasses comprise a camera module, a driving mechanism and a motion sensor, the driving mechanism is used for adjusting the posture of the camera module, and the motion sensor is used for monitoring head posture data of a user; the method comprises the following steps: determining a photographing compensation angle of a camera module, wherein the photographing compensation angle is determined according to one of head posture data and an adjustment instruction sent by a control terminal; and controlling the driving mechanism to adjust the posture of the camera module according to the photographing compensation angle. By means of the mode, the camera shooting visual angle deviation caused by the head movement in the photographing picture previewing process of the user can be effectively counteracted, it is ensured that the camera shooting module always keeps the stable view finding angle, and therefore real-time view finding of the non-screen photographing glasses in the previewing stage is achieved; and the shooting experience, the shooting precision and the shooting efficiency of the user during use are improved.

The invention discloses an intelligent glasses system and method for health risk monitoring of old people, and belongs to the technical field of intelligent health monitoring. The system is integrated on a glasses body and comprises a multi-mode sensing module, an edge calculation module and an interaction module. The method comprises the following steps: synchronously acquiring physiological parameters and behavior posture data of a user, extracting head posture change characteristics for representing body instability on a local edge side, performing time sequence fusion on the head posture change characteristics and the physiological parameter data, and inputting the fusion result into a pre-trained risk identification model to identify a falling risk or a sudden disease risk. And early warning is triggered when the risk is identified. According to the invention, non-inductive, continuous and accurate monitoring and early warning of high-risk scenes of old people are realized, the defects of single data, high false alarm rate and large response delay of existing equipment are overcome, and the real-time performance and reliability of health monitoring are remarkably improved.

A method for head posture judgment is provided. A plurality of head images are received to obtain raw label data corresponding to the head images. The raw label data include Euler angles marked in order of the x-y-z directions. The raw label data are converted into updated label data. The updated label data include Euler angles marked in order of the y-x-z directions. The updated label data of the head images are input into a deep learning network model. A loss function is used to train the deep learning network model to obtain a head posture detection model. Calibration images and real-time images are input into the head posture detection model to determine head posture.

The invention provides a virtual human interaction action generation method for unmarked augmented reality equipment, and the method comprises the steps: obtaining a head posture, a hand posture and a first-person visual angle image sequence of a user through a built-in sensor of an augmented reality helmet, and jointly reconstructing a whole-body posture of the user on the basis; generating an interaction action of a virtual human according to the reconstructed user posture; according to the method, the interactive action of the virtual human can be more natural and real and is kept consistent with the actual behavior of the user; in the field of man-machine interaction, the prior art depends on user control and needs additional hardware equipment and manual participation to maintain an interaction process; or depending on predefined conditions, generalization and practicability are poor; compared with the prior art, additional manual intervention and predefined design are not needed, interaction with the natural response virtual human can be achieved only through the lightweight augmented reality helmet, hardware dependence and labor cost are reduced, automation and usability of the method are promoted, and interaction convenience and immersion are improved.

The invention is suitable for the field of sight line estimation, and discloses a 3D sight line target estimation method and device, and the method comprises the steps: obtaining an original image, carrying out the extraction of an original feature map, a head posture vector, and an eye feature vector, and obtaining a 3D sight line direction vector according to the head posture vector and the eye feature vector; constructing a view field graph corresponding to the original image based on the plane component of the 3D sight line direction vector; constructing a scene feature graph according to the view field graph and the original feature graph; integrating the scene feature map and the depth component of the 3D sight line direction vector based on a ViT model, and generating an updated scene mark and an updated sight line mark; and inputting the updated sight line mark into a pre-trained dichotomy model to obtain a classification result, when the classification result is that the sight line target is in the image, generating a heat map according to the updated scene mark, and screening out a pixel point with the maximum heat value in the heat map as an estimated sight line target position, the precision and reliability of sight target estimation can be improved.