36 results about "Distracted driving" patented technology

The invention discloses a distraction driving detection model based on domain self-adaption, and aims to solve the problem that the performance of an existing distraction driving detection model is reduced due to the fact that distribution differences exist between a source domain and a target domain in the aspects of a driving environment, a shooting view angle, an illumination condition, driver characteristics and the like. According to the model, joint distribution alignment of feature and category prediction is realized through a label guiding domain adversarial network, and the adaptive capacity of the model in a cross-domain scene is remarkably improved. In order to further enhance the discrimination capability and generalization of the model, optimization is carried out from two levels of feature extraction and training strategy: a local feature enhancement module aims to improve the feature representation capability of key local behaviors; and according to the category structured CutMix strategy, the generalization ability of the model is enhanced by constructing a mixed sample. Experimental results show that the detection accuracy and robustness of the model under the cross-domain condition are effectively improved, a reliable technical scheme is provided for driver state monitoring, and the method has important application value for road traffic safety guarantee.

The invention discloses a truck driver safe driving evaluation method based on SEDEA, and the method comprises the steps: regarding the vehicle characteristics, behavior characteristics, road characteristics and environment characteristics of a vehicle in a driving process as external resources of a driver safe driving task, and regarding the characteristics of overspeed driving, fatigue driving, distraction driving and the like as risk results after the resources are utilized; the method comprises the following steps: taking resource utilization efficiency evaluation as a core means to measure the safe driving ability of a truck driver, and constructing a weight system containing multiple inputs and multiple outputs and a resource utilization efficiency evaluation model to calculate a safe driving score of the driver; and various safety indexes and quantitative classification management measures of the driver are given based on the model. The model is in close contact with the safe driving management situation of the truck driver, not only considers the resource input and the efficiency output of the safe travel task of the truck driver, but also sets the optimization direction between different inputs and outputs through a weight system.

A method (400) for distraction driver detection and alert includes receiving vehicle speed data (20) from one or more wheel speed sensors (122) disposed on a vehicle (105), and receiving proximity data from one or more proximity sensors disposed on the vehicle, the proximity data indicating a distance (32) of the vehicle relative to any object in front of the vehicle. The method also includes executing a distracted driver detection algorithm (300) that uses the vehicle speed data and the proximity data to determine that the vehicle is interrupting the stop-time traffic flow and that a driver of the vehicle is distracted from operating the vehicle. The method includes instructing a system (270) of the vehicle to output an alert to reengage a driver based on determining that the vehicle is interrupting the stop-time traffic flow and the driver of the vehicle is distracted.

The present invention aims to provide a storage unit that can be attached to a bicycle handlebar, allowing for the storage of a smartphone while simultaneously preventing distracted driving. [Structure] To achieve the above objectives, the present invention: Firstly, a storage unit for bicycle handlebars is attached to the handlebars at the center of the handlebars by fastening means such as screws at the mounting point, with a storage compartment protruding from the mounting point toward the rider, the storage compartment being opened and closed with a lid, and the horizontally elongated storage compartment being a storage space for storing a smartphone horizontally. The above is my proposal.

The invention discloses wearable interaction hardware equipment, and relates to the technical field of man-machine interaction. S1, equipment wearing and setting; s2, reporting contact information; s3, automatically replying the voice message; s4, performing voice reply by the user; and S5, traffic light pause state notification. According to the wearable interaction hardware equipment, a user does not need to hold a mobile phone to check or edit messages, distracted driving is avoided through voice interaction of a Bluetooth headset, the traffic accident risk is reduced, meanwhile, road congestion caused by distraction of a driver is reduced, and timely transmission of emergency messages and delayed processing of non-emergency messages are ensured through emergency feature word recognition; not only are important cooperation or help information not missed, but also normal driving is not interfered, the non-emergency message is automatically replied and the optimal contact time is informed, the communication cost of a user is reduced, and the non-emergency message is broadcasted and replied on the premise of ensuring sufficient time in combination with a traffic light pause state, so that the message processing efficiency is improved, and message accumulation is avoided.

Raw sensor data from a device of a user is collected. The raw data does not include location data for the device. The data is preprocessed to identify driving trips in which the user had the device. The trip data is provided to machine-learning models (MLMs) as input and the MLMs provide as output predictions for each trip's estimated speeds, estimated number of hard brakes and a degree of each hard brake, and estimated degrees of distracted driving. A duration, time of day, and calendar date of each trip is identified. The predictions, duration, time of day, and calendar date are modified into four factor values. The four modified factor values are combined into an overall driver characteristic value and provided to a network service associated with the user for purposes of providing or modifying services provided to the user through the network service based on the overall characteristic value.

The invention discloses an emotion intervention method and device, a vehicle and a storage medium. The method comprises the steps that video data and audio data of a driver and passengers in the vehicle driving process are acquired; performing emotion state detection on the driver according to the video data and the audio data to obtain an emotion score of the driver; when it is determined that the driver and the passenger are in a negative emotional state according to the emotional score, an intervention strategy is determined according to the emotional score, and the intervention strategy is used for smoothing the negative emotion of the driver and the passenger; and performing emotion intervention on the driver and passengers according to the intervention strategy. According to the method, when it is determined that the driver is in the negative emotional state based on the detected emotional score of the driver in the vehicle driving process, emotional intervention is conducted on the driver according to the emotional score so as to flatten the negative emotions of the driver, the situation that the driver is distracted to drive, and consequently the driving risk of vehicle driving is increased can be avoided, and the driving safety is improved. And the driving safety of vehicle driving can be improved.

The invention discloses a safe driving control method based on steering wheel grip strength monitoring, which comprises the following steps: S1, acquiring grip strength data such as grip strength values and distribution areas in real time through array type pressure sensor groups arranged in the circumferential direction of a steering wheel; s2, preprocessing the data and extracting characteristic parameters such as an average grip strength value and a distribution balance degree; s3, inputting the characteristic parameters into a pre-trained deep learning model, and identifying three types of driving states including fatigue, distraction and sudden health abnormality; s4, triggering a corresponding strategy according to an identification result: starting graded early warning and adjusting vehicle power parameters in fatigue or distraction driving, starting emergency control in case of abnormal health, and automatically executing speed reduction, lane keeping and alarming; and S5, reserving a 3s driver feedback window period, judging whether to release the control strategy based on the recovery coefficient, and if no effective feedback is received, continuously executing until the vehicle stops. The grip strength characteristic can be accurately monitored, the abnormal state can be recognized, and the driving safety can be effectively improved by being linked with a vehicle active control system.

The application relates to a light-weight distraction driving recognition method and system based on an improved ConvNeXt, and relates to the field of image classification of deep learning. The method comprises the following steps: 1. obtaining a driving scene image and performing pretreatment; 2. constructing an improved ConvNeXt model, and replacing a ConvNeXt Block module in an original ConvNeXt model with a CR-Former module; 3. iteratively training to obtain a distraction driving behavior recognition model; and 4. inputting an image into the distraction driving behavior recognition model to output a driving behavior recognition result. The system comprises an image acquisition and processing module, a model construction and configuration module, a model iterative training module and a behavior recognition output module which are connected in sequence. The application has the advantages that the model calculation cost is significantly reduced, the hardware resource occupation is reduced, and the capturing capability of the model for driving behavior detail features is effectively improved.

To more accurately detect the driver's condition. [Solution] The driver state detection device of the embodiment includes a recognition unit that recognizes at least one of the gaze or face direction of the driver of a moving object, and a determination unit that determines whether the driver is driving while distracted based on the recognition result of the recognition unit and a distraction determination condition. The determination unit makes a determination of distracted driving using a first distraction determination condition when the turning of the moving object is not predicted or the moving object is not turning, and makes a determination of distracted driving using a second distraction determination condition when the turning of the moving object is predicted or the moving object is turning. When the determination unit determines whether the moving object is turning, it changes the turning determination condition for determining whether the moving object is turning depending on whether the turning direction of the moving object is left or right.

Apparatuses, systems and methods are provided for determining vehicle driver distractions. More particularly, apparatuses, systems and methods are provided for receiving, via a vehicle interior data receiving module, vehicle interior data from at least one vehicle interior sensor, wherein the vehicle interior data is representative of an initial position of a body of at least one vehicle occupant; tracking, via a body tracking module, changes in the body of the at least one vehicle occupant from the initial position of the body; and predicting, via a driver action prediction module, one or more driver actions based at least on the changes in the body of the at least one vehicle occupant.

The invention provides a safe driving monitoring method based on multi-modal data, and the method comprises the following steps: S1, collecting video information during driving through a camera, separating audio information from the video information, and converting the audio information into text information through a voice recognition model; s2, inputting the video information into a target detection model, and when the driver is recognized to be in a fatigue or distracted driving state, generating a corresponding prevention measure; s3, synchronously performing the step S2, respectively extracting a visual initial feature, an audio initial feature and a text initial feature from the video information, the audio information and the text information, inputting the visual initial feature, the audio initial feature and the text initial feature into a multi-modal sentiment analysis model based on a mixed attention mechanism, and when the driver is recognized to be in an emotional driving state, executing the step S3; and corresponding prevention measures are generated. According to the invention, double-line detection is adopted to identify whether the driver is in fatigue, distraction or emotional driving, so that safe driving monitoring is more reliable and effective.

The application discloses a method for recognizing distraction driving behavior based on an improved DCH neural network, which comprises the following steps: constructing a hyper-dimensional calculation model; firstly, constructing an improved DCH neural network; replacing an ADAM optimizer in the DCH model with an improved artificial traveling mouse optimizer; then, using the improved DCH neural network to perform hyper-dimensional vector coding; in a hash code space, the Hamming distance of hash codes of semantically similar images is close, and the Hamming distance of semantically dissimilar images is far; finally, classifying image information of the same category to form a category hyper-dimensional vector; constructing an AdaptHD retraining model; using the model to calculate a classification accuracy rate, and calculating the similarity between all training samples and the category hyper-dimensional vector again; training the model, testing a data set by using the trained model; and outputting a test result; and the application improves the efficiency of the algorithm within an acceptable range of detection precision, reduces the time for training the model, and no longer depends on high-performance edge computing devices.

A distracted driving analysis system for identifying distracted driving events is provided. The system includes a processor in communication with a memory device programmed to: (i) receive driving event records, each driving event record including phone usage by a user, wherein a driving event record is labeled as an actual distracted driving event or a passenger event, (ii) divide the driving event records into at least two clusters based at least in part upon common features and the labels of each driving event record by processing the plurality of driving event records with a semi-supervised machine learning algorithm, (iii) generate a trained model based at least in part upon the at least two clusters, (iv) process a new driving event using the trained model, (v) assign the new driving event to one of the clusters using the trained model, and / or (vi) determine whether the new driving event is an actual distracted driving event or a passenger event.

Systems and methods are provided for programmatically verifying an origin of abnormal driving. Some examples of abnormal driving may include aggressive driving (e.g., tailgating, cut-in lane, etc.), distracted driving (e.g., swerving, delayed reaction, etc.), and reckless driving (e.g., green light running, lane change without signaling, etc.). For example, the systems and methods may receive an identification of a second vehicle performing abnormal driving from an ego vehicle; initiate a verification process of the identification of the abnormal driving; access driving data associated with an origin of the abnormal driving, wherein the driving data includes the ego vehicle and the second vehicle; using the driving data, determine a confirm or deny decision regarding the identification of the second vehicle from the ego vehicle; and provide the confirm or deny decision.

The invention discloses a detection and intervention system and method for distraction driving of a driver and a vehicle, and belongs to the technical field of vehicle detection. The detection and intervention system for distraction driving of the driver comprises a control module, and an execution module, an infrared image detection unit, a steering wheel pressure array detection unit and a positioning detection unit which are respectively connected with the control module, the control module is used for determining a comprehensive distraction score according to the head posture video picture data, the trunk offset data, the pressure distribution data and a preset environment compensation coefficient; under the condition that the driver is determined to have driving distraction according to the comprehensive distraction score, determining a distraction grade of driving distraction and a corresponding target intervention measure according to the comprehensive distraction score, a preset distraction grade interval and a corresponding relation between the preset distraction grade and the intervention measure, and sending the distraction grade and the corresponding target intervention measure to an execution module; the execution module is configured to execute the target intervention measure. The driving behavior of the driver can be quantitatively evaluated and graded intervention can be performed.

The invention provides a forklift-based driver state monitoring method and a forklift-based driver state monitoring system. The forklift-based driver state monitoring system comprises a multispectral imaging unit, a processor and an alarm unit, according to the method, RGB and infrared images are synchronously acquired, and a fusion weight is adaptively calculated based on a brightness mean value and an infrared signal-to-noise ratio, so that multispectral dynamic fusion is realized; recognizing targets such as human faces, cigarettes, mobile phones, transparent glasses, sunglasses and the like by using an improved lightweight detection network, and performing behavior judgment in combination with image-level and video-level classification networks; an MC Dropout uncertainty evaluation and consistency window fusion strategy is adopted, and events such as smoking, calling and distracted driving are output only when two branches are judged to be consistent and the reliability meets a threshold value. The system can realize real-time monitoring on low-computing-power edge equipment, has high stability and low false alarm rate in strong light, night, shielding and vibration scenes, and remarkably improves the safety of the working environment of the forklift.

Raw sensor data from a device of a user is collected. The raw data does not include location data for the device. The data is preprocessed to identify driving trips in which the user had the device. The trip data is provided to machine-learning models (MLMs) as input and the MLMs provide as output predictions for each trip's estimated speeds, estimated number of hard brakes and a degree of each hard brake, and estimated degrees of distracted driving. A duration, time of day, and calendar date of each trip is identified. The predictions, duration, time of day, and calendar date are modified into four factor values. The four modified factor values are combined into an overall driver characteristic value and provided to a network service associated with the user for purposes of providing or modifying services provided to the user through the network service based on the overall characteristic value.

The invention relates to the technical field of automobile control, in particular to a vehicle-mounted multi-mode navigation interaction method, system and equipment and a storage medium. The vehicle-mounted multi-mode navigation system comprises: a navigation prompt module, which is used for outputting a driving navigation prompt; the safety early warning module is used for outputting a driving safety early warning prompt; the head display module is used for projecting a navigation picture or a safety early warning picture on a vehicle window; the steering wheel tactile feedback module is used for providing navigation prompt vibration feedback or safety early warning vibration feedback on a steering wheel; and the controller is used for adjusting the working modes of the head display module and the steering wheel tactile feedback module according to the driving navigation prompt and the driving safety early warning prompt. According to the embodiment of the invention, by constructing a multi-mode navigation interaction system framework and integrating visual and tactile information transmission channels, the dependence of a driver on a central control screen is remarkably reduced, the sight line transfer frequency is reduced, the distraction driving problem caused by a traditional navigation system is effectively relieved, and the driving safety is improved.

This disclosure presents a driving behavior detection method and a training method for a driving behavior detection system. The method includes: acquiring an image dataset of the user to be detected; acquiring a driving behavior detection type; determining the data to be detected based on the image dataset and the driving behavior detection type; inputting the data to be detected into a joint multi-task driving behavior detection system, where the corresponding model performs behavior detection and outputs the driving behavior detection results for the user. This disclosure enables efficient and accurate joint multi-task detection of two typical non-standard driving behaviors: distracted driving and fatigued driving, laying a solid foundation for improving driving safety. Furthermore, it supports single-task detection, either focusing solely on distracted driving or fatigued driving, improving the flexibility and intelligence of the driving behavior detection process.

The invention discloses a grading intervention method for distracted driving, a vehicle machine and a program product, and relates to the technical field of intelligent cabins, and the method comprises the steps: collecting visual data, auditory data and tactile data in a cabin; performing feature extraction on the collected visual data, auditory data and tactile data by using different feature extraction models; performing weighted average on the extracted visual features, auditory features and tactile features to obtain a distraction value of the driver in the cabin; and performing grading intervention on the distraction state of the driver according to the value interval of the distraction value. According to the grading intervention method for distraction driving, the vehicle machine and the program product, the false alarm rate of distraction early warning can be effectively reduced.

A computer system is provided and is programmed to: (1) receive sensor data associated with a primary vehicle; (2) determine a current condition of a driver of the primary vehicle based upon the sensor data of the primary vehicle; (3) determine a threat level for the primary vehicle based upon the current condition of the driver of the primary vehicle and the sensor data of the primary vehicle; (4) activate at least one action in the primary vehicle based upon the threat level for the primary vehicle; and / or (5) electronically transmit the threat level of the primary vehicle to one or more additional vehicles.

A device housing system contained within a vehicle is provided herein, the device housing system including a front surface, a back surface, a first side member, a second side member, a top member, a bottom member, and a control module configured to communicate with a motion detection mechanism of the vehicle and with an internet-enabled device. The front surface, the first side member, the second side member, and the bottom member define a placement slot for the internet-enabled device, the placement slot being configured to support the internet-enabled device. The device housing system prevents a user from operating an internet-enabled device when the vehicle is in motion, and thus, limits distracted driving. Accordingly, the device housing system increases passenger and vehicle safety.

This prevents excessive notifications about distracted driving from being sent to the driver. [Solution] The vehicle 100 is configured to allow the driver to select one of several driving modes, including a first driving mode for normal driving and a second driving mode for driving on rough roads. The control device 6 of the vehicle 100 monitors the driver of the vehicle 100 and notifies the driver of distracted driving. When the driving mode is the second driving mode, the control device 6 is configured to suppress notifications regarding distracted driving compared to when the driving mode is the first driving mode.

This helps prevent delays in cargo handling operations while also preventing accidents caused by distracted driving. [Solution] The distraction monitoring system S comprises a shooting unit 12, an insertion / removal detection unit 22, a gaze estimation unit 24, a distraction determination unit 28, and a speed change unit 30. The shooting unit 12 photographs the forks of the forklift to generate a fork image and also photographs the driver to generate a head image. The insertion / removal detection unit 22 detects the insertion and removal of the forks from the pallet based on the fork image, and the gaze estimation unit 24 estimates the direction of the driver's gaze based on the head image. The distraction determination unit 28 determines whether the driver is distracted while inserting or removing the forks, based on the estimated direction of the driver's gaze during the insertion and removal of the forks. If the distraction determination unit 28 determines that the driver is distracted, the speed change unit 30 changes the operating speed of the forklift.

The invention discloses a distracted driving behavior classification method based on improved GRU optimization super-dimensional calculation post-processing, and the method comprises the steps: obtaining an image data set for recognizing the distracted behavior of a driver, and carrying out the preprocessing; building a super-dimensional calculation model, and adjusting the image data set into a training super-dimensional vector through convolution and pooling coding to obtain a category super-dimensional vector; performing AdaptHD retraining on the training samples, removing the training samples subjected to misclassification from wrong categories, and adding the training samples to correct categories again; and constructing a GRU model, replacing an ADAM optimizer in the GRU model with a chaos evolution optimizer, and carrying out further post-processing classification by taking the Hamming distance or cosine similarity between a training sample input in an AdaptHD retraining process and each category of super-dimensional vector as input to obtain a final classification result. Compared with the prior art, the GRU is integrated in the post-processing stage of the super-dimensional calculation process, and efficient and accurate behavior recognition is achieved.

The application relates to the technical field of data processing, and discloses a distraction driving behavior detection method based on a neural network model and related devices, which comprises the following steps: obtaining a distraction driving behavior image and a driving cabin image; performing label processing on the distraction driving behavior in the distraction driving behavior image to obtain a distraction driving behavior dataset; improving an original detection model to obtain an initial distraction driving behavior detection model; dividing the distraction driving behavior dataset into a training set and a verification set according to a preset proportion, training the initial distraction driving behavior detection model, and obtaining a target distraction driving behavior detection model; performing pretreatment on the driving cabin image to obtain a distraction detection image; and performing distraction driving behavior detection on the distraction detection image through the target distraction driving behavior detection model to obtain a distraction driving behavior detection result, which can effectively improve the accuracy of the detection result.

This invention discloses a method for recognizing distracted driving behavior based on contrastive learning and real-time edge computing, relating to the field of intelligent traffic monitoring technology. The method includes: acquiring a target dataset; constructing a YOLOv5 detection model and pre-training it; converting the pre-trained YOLOv5 detection model into a YOLOv5 classification model and retraining it on the target dataset to obtain a trained PT model; improving the PT model based on contrastive learning; pruning the improved PT model using Batch Normalization (BN) layers, followed by quantization; converting the PT model to an ONNX model and then to an RKNN model and pre-compiling it; finally, using the pre-compiled RKNN model for real-time inference on an edge computing platform. This invention utilizes contrastive learning to enable the model to learn a feature extractor with representational capabilities, effectively improving the accuracy of driver behavior classification.

The application provides a distraction driving detection method based on improved YOLOv5, which detects key points by analyzing the head posture and body posture of the driver. First, an AIFI module is introduced based on YOLOv5. The module can better capture subtle changes and fine features of the driver's behavior by optimizing the interaction between deep layers of the same feature map, avoiding the complex overall performance evaluation and data analysis process in the SPPF module, and thus being more efficient in calculation and suitable for real-time application scenarios. Second, the original Concat module of the neck layer is replaced with a TFE module, which fuses feature maps through adaptive scale adjustment and channel attention weighting, thereby improving the accuracy of small target detection such as mobile phones and cups. Finally, a shape-IoU module that considers shape information is introduced, which enhances the robustness of the target with occlusion, deformation or partial occlusion, and provides more stable performance compared to the original IoU.

A computer system is provided and is programmed to: (1) receive sensor data associated with a primary vehicle; (2) determine a current condition of a driver of the primary vehicle based upon the sensor data of the primary vehicle; (3) determine a threat level for the primary vehicle based upon the current condition of the driver of the primary vehicle and the sensor data of the primary vehicle; (4) activate at least one action in the primary vehicle based upon the threat level for the primary vehicle; and / or (5) electronically transmit the threat level of the primary vehicle to one or more additional vehicles.