Embodiments of the disclosure provide a warning method and apparatus for a driving risk, a computing device and a storage medium, and the method includes: obtaining dangerous driving behavior data of a driver in a first time period, and obtaining a correspondence between a quantity of occurrences of dangerous driving behaviors of one or more drivers and a quantity of an actual occurrence of dangerous scenarios to the one or more drivers while driving; predicting, based on a quantity of actual occurrences of the dangerous driving behaviors of the driver, indicated in the dangerous driving behavior data of the driver, and the correspondence, a target quantity of times the driver is predicted to encounter one or more dangerous scenarios in the first time period; and generating warning information based on the target quantity of times.

The present disclosure relates to methods and apparatuses for identifying an object of interest of a user. One example method includes obtaining information about a line-of-sight-gazed region of the user and an environment image corresponding to the user, obtaining information about a first gaze region of the user in the environment image based on the environment image, where the first gaze region is used to indicate a sensitive region determined by using a physical feature of a human body, and obtaining a target gaze region of the user based on the information about the line-of-sight-gazed region and the information about the first gaze region. The gaze region is used to indicate a region in which a target object gazed by the user in the environment image is located.

An advice target location at which a user had a predetermined emotion, for example, is determined based on location information, user biological information, and user transportation means information, which have been acquired by a terminal device (20) being used by the user. Advice information containing information indicating an advice presentation region set by a server device (50) is generated based on the advice target location. This advice information is supplied from the server device (50) to the terminal device (20), so that the terminal device (20) presents advice. With this, advice as to locations pedestrians find dangerous can be presented to drivers, and advice as to locations drivers find dangerous can be presented to pedestrians. Accordingly, accidents and the like can be prevented.

Systems and methods for providing continuous safe-driver training safely are provided. A safe-driving challenge may be presented to an operator of a vehicle. Data captured by sensors associated with the vehicle may be analyzed to determine whether the operator of the vehicle has completed the safe-driving challenge. Based on a determination that the operator of the vehicle has completed the safe-driving challenge, a notification may be provided to the operator (e.g., indicating to the operator that he or she has successfully completed the challenge). A processor may randomly select whether a reward is to be provided to the operator of the vehicle based on the determination that the operator of the vehicle has completed the safe-driving challenge. Moreover, if a reward is to be provided to the operator of the vehicle, a processor may randomly select a type of reward to be provided to the operator of the vehicle.

In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

A method for driver identification including recording a first image of a vehicle driver; extracting a set of values for a set of facial features of the vehicle driver from the first image; determining a filtering parameter; selecting a cluster of driver identifiers from a set of clusters, based on the filtering parameter; computing a probability that the set of values is associated with each driver identifier of the cluster; determining, at the vehicle sensor system, driving characterization data for the driving session; and in response to the computed probability exceeding a first threshold probability: determining that the new set of values corresponds to one driver identifier within the selected cluster, and associating the driving characterization data with the one driver identifier.

Disclosed are a fatigue state detection method and apparatus, a medium and a device. The method includes: obtaining image blocks containing an organ area of a target object from a plurality of video frames collected by a camera apparatus disposed in a mobile device, to obtain an image-block sequence that is based on the organ area; determining a fatigue state type of the target object based on the image-block sequence of the organ area; sending the image-block sequence to a cloud server if the fatigue state type meets a first preset type, and rendering the cloud server to detect a fatigue level of the target object based on the image-block sequence; and receiving fatigue level information about the target object that is returned by the cloud server. The present disclosure may improve accuracy of fatigue state detection, thereby helping to improve driving safety of the mobile device.

A vehicular driver monitoring system includes a camera disposed within a vehicle and having a field of view interior of the vehicle, and an adjustable mirror disposed within the vehicle. The adjustable mirror includes a mirror reflector that reflects light incident thereat. The camera views the adjustable mirror and captures image data based on light reflected by the mirror reflector of the adjustable mirror toward the camera. With a driver of the vehicle sitting in a driver seat of the vehicle, light reflected off a portion of the driver is reflected by the mirror reflector of the adjustable mirror toward the camera. An electronic control unit (ECU), responsive to processing by an image processor of image data captured by the camera, adjusts the adjustable mirror to reflect the portion of the driver toward a region of the lens of the camera.

The present disclosure relates to a method and a device for recognizing speech in a vehicle. The method for recognizing the speech in the vehicle may include collecting one or more types of information, determining information to be linked with each other for speech recognition based on an information processing priority predefined corresponding to each type of the collected information, analyzing the determined information to perform the speech recognition for a signal input through a microphone, and extracting at least one of a wake up voice or a command voice through the speech recognition to control the vehicle. Therefore, the present disclosure has an advantage of more accurately performing the speech recognition by linking collected various information in the vehicle with each other.

A pupil detection device includes an eye area image obtaining unit that obtains image data representing an eye area image in a captured image obtained by a camera; a luminance gradient calculating unit that calculates luminance gradient vectors corresponding to respective individual image units in the eye area image, using the image data; an evaluation value calculating unit that calculates evaluation values corresponding to the respective individual image units, using the luminance gradient vectors; and a pupil location detecting unit that detects a pupil location in the eye area image, using the evaluation values.