Abstract: A method for recognizing product detection missed is performed by an electronic device. The method includes: obtaining a surveillance video collected by a surveillance device on a product production line; generating product detection results by inputting video frames of the surveillance video into a product detection model; generating detecting action recognition results by inputting the video frames of the surveillance video into an action recognition model; and determining whether product detection is missed based on the product detection results and the detecting action recognition results.

Abstract: The present disclosure provides a method for graph computing, an electronic device and a non-transitory computer-readable storage medium. An execution engine for managing execution of the graph computing is configured in a CPU. One or more interface functions running in a GPU for processing parameters of the graph computing are configured. During the execution of the graph computing, the one or more interface functions are called by the execution engine through a graph computing interface. The interface functions are executed in the GPU in parallel through multiple parallel threads. The interface functions are configured to process a plurality of graph vertexes in parallel. The multiple parallel threads are configured to feedback respective execution results of the interface functions to the execution engine. The graph computing is completed by the execution engine according to the execution results.

Abstract: A method for training a trajectory planning model, an apparatus, and computer storage medium are provided. The method may include: obtaining an image of a physical environment in which a vehicle is located via at least one sensor of the vehicle, the image including multiple objects surrounding the vehicle; obtaining a feature chart indicating multiple initial trajectory points of the vehicle in the image from a trajectory planning model based on the image; identifying the image to determine in the image a first area associated with a road object in multiple objects and a second area associated with a non-road object in the multiple objects; determining a planning trajectory point based on positional relationship of the multiple initial trajectory points with respect to the first area and the second area; and training a trajectory planning model based on the planning track point and the actual trajectory point of the vehicle.

Abstract: A method for acquiring point of interest (POI) representation information, includes: acquiring a POI relationship graph, and acquiring a node relationship set of each node on the POI relationship graph based on the POI relationship graph, in which the POI relationship graph includes edges representing relationships between nodes of POIs, and the node relationship set includes relationship paths where the node reaches all second-order neighbor nodes via any one first-order neighbor node; acquiring a spatial distance matrix between the node and corresponding second-order neighbor nodes; and determining first POI representation information of the node based on the node relationship set of the node and the spatial distance matrix.

Abstract: An information acquisition method and apparatus, an electronic device, and a storage medium are suggested, which relates to the field of artificial intelligence (AI) such as machine learning and distributed storage. The method may include: determining hardware devices located on a to-be-processed node in response to determining that a trigger condition is met; selecting an x processing unit (XPU) from the hardware devices; and acquiring predetermined information of the XPU by using a system management interface (SMI) management tool corresponding to the XPU, the predetermined information including model information of the XPU.

Abstract: Provided are a method and apparatus for processing map information, a device, and a storage medium which relate to the field of computer technology and, in particular, to the fields of intelligent transportation and computer vision technology. The specific implementation includes that an object in an image for query is recognized to obtain an object recognition result and that a target point of interest matching the image for query is selected from at least one candidate point of interest of an electronic map according to the object recognition result.

Abstract: The present disclosure provides an ultrasonic radar array, an obstacle detection method and system. The method includes: obtaining obstacle information collected by ultrasonic radars in an ultrasonic radar array in an obstacle scenario; judging false detection and missed detection for the obstacle information collected by ultrasonic radars according a preset rule; processing the obstacle information collected by the ultrasonic radars according to the judgement result; determining a position of the obstacle according to the processed obstacle information collected by the ultrasonic radars.

Abstract: The present disclosure provides a method and system for processing an obstacle detection result of an ultrasonic sensor array. The method comprises: obtaining obstacle coordinates collected by ultrasonic sensors in an ultrasonic sensor array; inputting the obstacle coordinates collected by ultrasonic sensors in the ultrasonic sensor array into a pre-trained neural network model, to obtain true or false labels for the obstacle coordinates collected by the ultrasonic sensors output by the pre-trained neural network model; process the obstacle coordinates collected by the ultrasonic sensors in the ultrasonic sensor array according to the true or false labels. The present disclosure improves the accuracy in the obstacle detection of the ultrasonic sensor array, avoids detection error and omission and improves the driving safety.

Abstract: The present application provides a model training, image processing method, device, storage medium, and program product relating to deep learning technology, which are able to screen auxiliary image data with image data for learning a target task, and further fuse the target image data and the auxiliary image data, so as to train a built and to-be-trained model with the fusion-processed fused image data. This implementation can increase the amount of data for training the model, and the data for training the model is determined is based on the target image data, which is suitable for learning the target task. Therefore, the solution provided by the present application can train an accurate target model even if the amount of target image data is not sufficient.

Abstract: The present application provides an image processing method, an image processing apparatus, an electronic device and a computer-readable storage medium, and relates to the field of image processing technologies. An implementation includes: acquiring an image to be processed; converting the image to be processed into a three-channel YUV image; performing a convolution operation on a Y-channel image, a U-channel image and a V-channel image in the three-channel YUV image to generate an R-channel image, a G-channel image and a B-channel image, respectively, and acquiring a three-channel RGB image; and pre-processing the three-channel RGB image. According to the present application, the image pre-processing speed can be improved.

Abstract: The present disclosure provides an event extraction method, an event extraction device and an electronic device, and it relates to the field of computer data processing, in particular to the field of knowledge graph. The event extraction method includes: acquiring text information; determining a plurality of pieces of question information ranked in a sequential order in accordance with the text information; and inputting vector information for each piece of question information into an extraction model in accordance with the sequential order to acquire extraction information for each piece of question information.

Abstract: The present disclosure provides a method and apparatus of editing a multimedia file, an electronic device, and a storage medium. The method includes: acquiring a multimedia file and an operation information contained in each editing track of a plurality of editing tracks, wherein the operation information is acquired to edit the multimedia file; detecting a multimedia file frame contained in the each editing track at a current time in a case of editing the multimedia file in a chronological order; editing the multimedia file frame contained in the each editing track at the current time by using the operation information associated with the multimedia file frame, so as to obtain an edited file frame contained in the each editing track at the current time; and performing an encoding synthesis on the edited file frame contained in the each editing track at the current time.

Abstract: This disclosure provides a video generation method, a video generation apparatus, an electronic device, a storage medium and a program product, and relates to the field of artificial intelligence technology, and in particular to the field of computer vision technology and deep learning technology. A specific implementation includes: obtaining document content information of a document; extracting, from the document content information, populating information for multiple scenes in a preset video template; populating the populating information for the multiple scenes into corresponding scenes in the preset video template, respectively, to obtain image information of the multiple scenes; generating audio information of the multiple scenes according to the populating information for the multiple scenes; generating a video of the document based on the image information and audio information of the multiple scenes.

Abstract: A method for text generation, relates to a field of natural language processing, including: obtaining corpus data; labeling the corpus data to obtain a first constraint element; obtaining a first generation target; and generating a first text matching the first generation target by inputting the corpus data and the first constraint element into a generation model.

Abstract: A multi-modal pre-training model acquisition method, an electronic device and a storage medium, which relate to the fields of deep learning and natural language processing, are disclosed. The method may include: determining, for each image-text pair as training data, to-be-processed fine-grained semantic word in the text; masking the to-be-processed fine-grained semantic words; and training the multi-modal pre-training model using the training data with the fine-grained semantic words masked.

Abstract: Provided are a method and apparatus for evaluating image relative definition, a device and a medium, relating to technologies such as computer vision, deep learning and intelligent medical. A specific implementation solution is: extracting a multi-scale feature of each image in an image set, where the multi-scale feature is used for representing definition features of objects having different sizes in an image; and scoring relative definition of each image in the image set according to the multi-scale feature by using a relative definition scoring model pre-trained, where the purpose for training the relative definition scoring model is to learn a feature related to image definition in the multi-scale feature.

Abstract: A method for operating a quantum computer is provided. The method includes: obtaining a quantum gate parameter of each quantum gate in a quantum circuit to be simulated; and generating a sub-measurement mode equivalent to the quantum gate based on the quantum gate parameter. The method further includes combining the sub-measurement mode equivalent to each quantum gate to obtain an overall measurement mode equivalent to the quantum circuit to be simulated as a whole; and sorting an operation order of the operation commands of each sub-measurement mode in the overall measurement mode, thereby obtaining the sorted overall measurement mode as a simulation result. This solution transforms the quantum circuit used for quantum computation into an equivalent measurement mode according to a one-way quantum computer computation model. Direct simulation computation on the quantum circuit can be avoided, and accordingly a simulation operation amount of a classical computer is greatly reduced.

Abstract: The present disclosure provides a method of training a depth estimation network, which relates to fields of computer vision, deep learning, and image processing technology. The method includes: performing a depth estimation on an original image by using a depth estimation network, so as to obtain a depth image for the original image; removing a moving object from the original image so as to obtain a preprocessed image for the original image; estimating a pose based on the original image and modifying the pose based on the preprocessed image; and adjusting parameters of the depth estimation network according to the original image, the depth image and the pose modified. The present disclosure further provides an apparatus of training a depth estimation network, a method and apparatus of estimating a depth of an image, an electronic device, and a storage medium.

Abstract: Embodiments of the present disclosure disclose a method and apparatus for determining a target anchor, a device and a storage medium. The method may include: extracting a plurality of feature maps of an original image using a feature extraction network; inputting the plurality of feature maps into a feature pyramid network to perform feature fusion, to obtain a plurality of fused feature maps; and using a region proposal network to implement operations as follows: determining an initial anchor of a network head using the fused feature map, based on a size of each fused feature map, and determining an offset parameter of the initial anchor, based on a ratio of the size of the fused feature map to the original image, and generating a plurality of candidate anchors in different directions, based on the offset parameter of the initial anchor.

Abstract: The present application discloses a method and apparatus for training a semantic representation model, a device and a computer storage medium, which relates to the field of natural language processing technologies in artificial intelligence.