Abstract: A method, device, computer equipment and storage medium for GNSS time synchronization are disclosed. The method includes: receiving data packet of NMEA protocol, reading a valid UTC time from data packet of NMEA protocol, and storing the read valid UTC time in a time synchronization controller; receiving PPS signal, capturing a local time output by local clock when PPS signal is generated, and storing the local time in the time synchronization controller; reading a last local time stored before the current local time and reading the stored latest UTC time as a UTC time corresponding to the last local time when the time synchronization controller receives the current local time; and determining, by the time synchronization controller, a local time correction amount according to the last local time and the UTC time corresponding to the last local time, and correcting the local clock according to the local time correction amount.

Abstract: A method and an apparatus for generating a three-dimensional point cloud image, a computer device and a storage medium. The method comprises: detecting three-dimensional shape coordinates of a three-dimensional object in an infrared camera coordinate system; detecting three-dimensional marker coordinates of the marker set in a marker measurement coordinate system at the same time; obtaining, according to a pre-stored first conversion relationship and a second conversion relationship parameter, three-dimensional color coordinates in the infrared camera coordinate system on the basis of the three-dimensional marker coordinates in the marker measurement coordinate system; and generating a three-dimensional point cloud image of the three-dimensional object in the infrared camera coordinate system by combining the three-dimensional shape coordinates and the three-dimensional color coordinates of the three-dimensional object in the infrared camera coordinate system.

Abstract: A method performed by an electronic device with one or more processors and memory includes receiving a plurality of frames of an image; selecting one frame of the plurality of frames as a main frame thereby leaving the rest of the plurality of frames as reference frames; aligning the reference frames with the main frame; determining fusion weights for a respective reference frame of the reference frames by comparing the main frame and the respective reference frame; and obtaining a weighted combination of the main frame and the reference frames based on the fusion weights. An electronic device configured for performing the method and a computer readable storage medium storing instructions for performing the method are also disclosed.

Abstract: A safety belt detection method, apparatus, computer device and computer readable storage medium are disclosed. The safety belt detection method includes the steps as follows. An image to be detected is obtained. The image to be detected is inputted into a detection network which includes a global dichotomous branch network and a grid classification branch network. A dichotomous result, which indicates whether a driver is wearing a safety belt and is output from the global dichotomous branch network, is obtained. A grid classification diagram, which indicates a position information of the safety belt and is output from the grid classification branch network, is obtained based on image classification. A detection result of the safety belt, indicating whether the driver is wearing the safety belt normatively, is obtained based on the dichotomous result and the grid classification diagram.

Abstract: The disclosure provides a processing system which comprises a plurality of processors independently executing instructions; a plurality of registers arranged in a manner corresponding to the plurality of processors. Each of the plurality of registers is configured to set first register bits for other processors except the corresponding processor, and the other processors can write the first register bits set for them to indicate event requests. And each processor of the plurality of processors is configured to learn the event requests from the remaining processors for it by reading the first register bits of the corresponding register. Therefore, the application establishes an inter-processor communication module for a plurality of processors in a processing system.

Abstract: A method of human pose estimation, including, receiving an image frame, extracting in at least a first stage a first stage image feature based on the image frame, extracting in at least a second stage a second stage image feature based on the first stage image feature, extracting in at least a subsequent stage a subsequent stage image feature based on the second stage image feature, up-sampling the subsequent stage image feature, up-sampling the second stage image feature, concatenating the first stage image feature, the up-sampled second stage image feature and the up-sampled subsequent stage image feature and outputting a feature map based on an output of the concatenation.

Abstract: A video encoding/decoding method and apparatus, a computer device, and a storage medium. The method comprises: for each non-key frame in a video frame sequence, dividing a current non-key frame into a plurality of sub-image blocks according to information of an object in the current non-key image, and determining an importance level of each sub-image block; according to a pre-stored first correlation between different importance levels and different resolutions, performing conversion to make each sub-image block in each non-key frame have a resolution corresponding to the importance level of the sub-image block, wherein in the first correlation, a higher importance level corresponds to a higher resolution, and the highest importance level corresponds to a target highest resolution; and performing video encoding on the video frame sequence to obtain encoded video data. By means of the method, memory and bandwidth occupied by a video can be reduced.

Abstract: The present invention discloses an intelligent cloud-based photo enhancement system to improve the quality and aesthetics of an image captured via an electronic device. The cloud-based photo enhancement system comprises a cloud database containing a collection of high-quality photos taken at various locations worldwide and under different environment settings. Information from photos stored in the cloud database taken using better photographic capability cameras is used to improve details that the new photo fails to capture. The system relies not only on machine learning models but also utilizes data associated with the new image collected from the hardware of the electronic device such as camera settings, GPS and phone sensors, which results in more reliable information.

Abstract: The application relates to a method, a device and a storage medium for storing and playing back unmanned driving information, wherein the unmanned driving information storage method comprises the following steps of: acquiring a plurality of data signals associated with an abnormal event when the abnormal event of an unmanned vehicle is monitored; calculating a first hash value for each data signal; storing each first hash value and the corresponding data signal as a key value pair in a key value database respectively; generating a first Merkel tree based on each first hash value; and storing the first Merkel tree and critical data signals of the plurality of data signals into a blockchain, and generating corresponding electronic money addresses and transaction timestamps. By this method, the correctness, integrity, fairness, and credibility of abnormal field data can be guaranteed, and too much storage space will not be occupied at the same time.

Abstract: The present invention discloses a liveliness detection technique. The technique is described for identifying facial attributes. The technique identifies the presented face in the image as real or deceptive. The system and method includes identifying the facial attributes and utilizing a multi task learning network. The neural network includes segmentation and classification functionalities. The final output is used to get pixel level semantic information and high level semantic information.

Abstract: The disclosure provides a bus-based communication system, which comprises a bus controller, a first buffer, a bus redundant controller, a second buffer and a comparator. The bus controller is configured to receive a bus input signal and send a bus output signal. The first buffer is connected with the bus controller in parallel to receive the bus input signal, and is configured to output the bus input signal after a preset time, wherein the preset time is two or more duty cycles of the bus controller. The bus redundant controller is connected to an output terminal of the first buffer and configured to receive a bus output signal delayed by at least T cycles output by the first buffer and output a bus redundant output signal.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a machine learning model configured to generate a predicted depth image, comprising receiving data representing training samples that include a plurality of image pairs, each image pair includes a target image and a reference image both capturing a particular scene from different orientations; for each of the plurality of image pairs, generating a compressed cost volume for the image pair; providing the compressed cost volume as an input to the machine learning model; generating, using the machine learning model, output data representing a predicted disparity map for the compressed cost volume; and generating a total loss using the predicted disparity map for the compressed cost volume, the total loss includes a boundary loss, an occlusion loss, and a transfer loss; and updating the plurality of parameters of the machine learning model by minimizing the total losses.

Abstract: Embodiments of this disclosure provide a system and method for reducing quantization errors at an output of a hardware device functioning as a neural network structure. During operation, the system can obtain type information associated with the neural network structure and construct a hardware model of the neural network structure based on the obtained type information, the hardware model comprising one or more paths for performing arithmetic operations. The system can formulate an optimization problem to reduce the quantization errors based on the constructed hardware model, the optimization problem being defined by an objective function and a set of constraints. The system can solve the optimization problem and configure the hardware device based on a solution to the optimization problem, thereby reducing the quantization errors at the output of the hardware device.

Abstract: One embodiment provides a system and method for generating vectorized maps. During operation, the system can obtain labeled training samples comprising images and vectorized ground truth information associated with the images. The system can generate, based on a machine learning model, a prediction output for each training sample, the prediction output comprising a vectorized map. The system can apply a differentiable rasterization technique to a vectorized map element within the vectorized map to obtain a segmentation mask, apply the differentiable rasterization technique to the vectorized ground truth information associated with the vectorized map element to obtain a ground truth segmentation mask, compute a segmentation loss function based on the segmentation mask and the ground truth mask, and update the machine learning model based on the segmentation loss function, thereby facilitating optimized generation of vectorized maps.

Abstract: An example fused convolutional layer, comprising, a comparator capable of reception of a first zero point and a multiply-accumulation result, a first multiplexer coupled to the comparator, wherein the first multiplexer receives a plurality of power-of-two exponent values, a shift normalizer, coupled to the first multiplexer, wherein the shift normalizer is capable of receiving the multiply-accumulation result and the plurality of power-of-two exponent values, wherein the shift normalizer limits a quantization of the multiply-accumulation result to a power-of-two scale and a second multiplexer coupled to an output of the shift normalizer, the first multiplexer and receives a second zero point and outputs an activation.

Abstract: A hot standby fault processing system is provided. The system includes a first detection layer configured to detect and record faults of vehicle configurations with low safety levels, a second detection layer configured to detect a fault of a vehicle configuration with critical safety level and send fault information, a fault collection layer configured to classify and record the received fault information, freeze the vehicle configuration corresponding to the fault information, and send the fault information to a higher processing layer, a first processing layer configured to receive fault information sent by the fault collection layer, process faults corresponding to the fault information according to a preset policy, and send fault information of faults that cannot be processed to system safety components, a redundancy part configured as a backup with the first processing layer and take over the operation of the first processing layer when the first processing layer fails.

Abstract: The present invention discloses a multi-frame image super resolution system that utilizes both deep learning models and traditional models of enhancing the resolution of an image so that minimal computational resources are used. A frame alignment module of the invention aligns the frames of the image after which a processing module configured within the system process the Y and the UV channels of the image by using multiple deep and traditional resolution enhancement models. A merging unit merges the output of the processors to produce a super resolution image incorporating the advantages of both of the image enhancement methods.

Abstract: The present invention provides an intelligent and a flexible video enhancement system with improved visual quality of the video for better user experience. The video enhancement system enhances various types of videos and provides a user with better visual experience based on video corresponding features and personal preferences. The video enhancement system includes a dynamic analysis and adaptive processing techniques for enhancing video quality for seamless user experience.

Abstract: A method, an apparatus, an electronic apparatus, and a storage medium for image distortion calibration are provided. In the method, an original image is obtained. The original image includes a background region and a foreground object region. A background distortion calibration is performed on the original image for obtaining a first calibrated image. The foreground object distortion calibration is performed to the original image for obtaining the second calibrated image. The background region in the first calibrated image and the foreground object region in the second calibrated image are fused for obtaining a calibrated image. The method can not only calibrate the distortion of the background of the original image but also preserve the shape of the foreground object, thus improving the presenting effect of the calibrated image.

Abstract: The present invention discloses a system and a method for providing multi-channel high-quality depth estimation from a monocular camera for providing augmented reality (AR) and virtual reality (VR) features to an image. The invention further includes the method to enhance generalization on deployment-friendly monocular depth inference pipeline with semantic information. Furthermore, a vivid and intact reconstruction is guaranteed by inpainting the missing depth and context within the single image input.