Abstract: A method of creating a delay queue architecture for relaying messages between a producer and a consumer is provided. The method includes creating an ordered sequence of delay queues. Each delay queue is assigned in the ordered sequence based on a fixed delay time associated with each delay queue. A first message is assigned to a first delay queue of the delay queues in the ordered sequence and a second message is assigned to a second delay queue of the delay queues. The first and second messages have time delays and are assigned to the first and second delay queues based on the time delays. A passage of a first-time delay is detected and the second message is moved to the first delay queue based on the passage of the first-time delay.

Abstract: A method of creating a delay queue architecture for relaying messages between a producer and a consumer is provided. The method includes creating an ordered sequence of delay queues. Each delay queue is assigned in the ordered sequence based on a fixed delay time associated with each delay queue. A first message is assigned to a first delay queue of the delay queues in the ordered sequence and a second message is assigned to a second delay queue of the delay queues. The first and second messages have time delays and are assigned to the first and second delay queues based on the time delays. A passage of a first-time delay is detected and the second message is moved to the first delay queue based on the passage of the first-time delay.

Abstract: A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.

Abstract: Provided is a fingerprint sensor, including: a substrate; a plurality of photosensitive devices disposed on the substrate; a light-emitting module disposed on a side, distal to the substrate, of the plurality of photosensitive devices; and a protective cover disposed on a side, distal to the substrate, of the light-emitting module, wherein the protective cover is provided with a plurality of conductive structures electrically connected to the light-emitting module, the plurality of conductive structures being in one-to-one correspondence with the plurality of photosensitive devices, and an orthographic projection of each conductive structure onto the substrate at least partially being overlapped with an orthographic projection of the photosensitive device corresponding to the conductive structure onto the substrate.

Abstract: A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.

Abstract: The disclosure provides a detection substrate and a detection device. The detection substrate includes: a base substrate including a photosensitive area and a peripheral area surrounding the photosensitive area; a plurality of organic photodetectors on the base substrate which are arranged in an array in the photosensitive area, the organic photodetectors each including a first electrode, an organic photoelectric detection function layer, and a second electrode which are stacked, and the second electrodes of all organic photodetectors being integrally arranged and extending from the photosensitive area to the peripheral area; and a bias line, the bias line being a strip-shaped line extending in a first direction in the peripheral area, the bias line being electrically connected to the second electrodes in the peripheral area, and the minimum distance between the bias line and the organic photoelectric detection function layers in the second direction is greater than a preset threshold.

Abstract: A method and system for positioning a target in a brain region are provided. The method includes: obtaining datasets of N persons at a first time point and a second time point after stroke; constructing a first lesion mapping functional network based on each resting-state functional magnetic resonance imaging image in a first stroke dataset; constructing an acute phase cognitive-lesion mapping functional network; constructing a chronic phase cognitive-lesion mapping functional network; comparing the acute phase cognitive-lesion mapping functional network with the chronic phase cognitive-lesion mapping functional network to obtain a key improvement network; calculating a whole-brain functional connectivity network with each voxel as a seed point, and performing spatial correlation calculation on the whole-brain functional connectivity network and the key improvement network to obtain a spatial correlation network; and determining a therapeutic target of the functional image to be positioned.

Abstract: Provided is a fingerprint sensor, including: a substrate; a plurality of photosensitive devices disposed on the substrate; a light-emitting module disposed on a side, distal to the substrate, of the plurality of photosensitive devices; and a protective cover disposed on a side, distal to the substrate, of the light-emitting module, wherein the protective cover is provided with a plurality of conductive structures electrically connected to the light-emitting module, the plurality of conductive structures being in one-to-one correspondence with the plurality of photosensitive devices, and an orthographic projection of each conductive structure onto the substrate at least partially being overlapped with an orthographic projection of the photosensitive device corresponding to the conductive structure onto the substrate.

Abstract: Disclosed are a detection substrate and a detection device. The detection substrate includes: a plurality of photoelectric converters arranged in an array, where the photoelectric converter includes a plurality of film layers, an area of an overlapping region between electrode film layers forming an internal capacitor in the photoelectric converter is less than an area of the other film layer in the photoelectric converter; and a drive circuit electrically connected to the photoelectric converters.

Abstract: A computer vision processor of a camera extracts attributes of persons or vehicles from hyperzooms generated from image frames. The hyperzooms represent traffic patterns. The extracting is performed using a feature extractor of an on-camera convolutional neural network (CNN) including an inverted residual structure. The attributes include at least colors of clothing of the persons or colors of the vehicles. Mobile semantic segmentation models of the CNN are generated using the hyperzooms and the attributes. Attribute analytics are generated by executing the mobile semantic segmentation models while obviating network usage by the camera. The attribute analytics are stored in a key-value database located on a memory card of the camera. A query is received from the server instance specifying one or more of the attributes. The attribute analytics are filtered using the one or more of the attributes to obtain a portion of the traffic patterns.

Abstract: A script-to-movie generation method for a computing device includes: obtaining a movie script; generating a list of actions according to the movie script; generating stage performance based on each action in the list of actions; extracting observation information from the stage performance; using a camera agent trained with a reinforcement learning algorithm to select a camera based on the observation information, where the camera includes camera setting that defines a position of the camera with respect to a character for which the camera shoots; using the selected camera to capture a video of the stage performance; and outputting the video.

Abstract: A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.

Abstract: A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.

Abstract: A method and system for extracting a multi-dimensional disconnection network region of symptom mapping: registering a lesion image to a brain standard space; obtaining diffusion magnetic resonance images and resting-state functional magnetic resonance images of healthy control groups; constructing a structural disconnection weighting network corresponding to lesions using a fiber tracking method according to the lesion image in the brain standard space and the diffusion magnetic resonance images; constructing a functional significant disconnection network corresponding to the lesions using a cross-correlation verification method according to the lesion image in the brain standard space and the resting-state functional magnetic resonance images; and determining the multi-dimensional disconnection network region of the lesions of symptom mapping according to the structural disconnection weighting network and the functional significant disconnection network, where the multi-dimensional disconnection network regi

Abstract: Provided are a preparation method for a tumor-infiltrating lymphocyte, and use thereof, particularly relating to a method for culturing a tumor-infiltrating lymphocyte, comprising: subjecting a TIL, which is derived from tumor tissues and not expanded in vitro, to in vitro expansion of at least one stage, and wherein the TIL is contacted with a CD28 agonist in the in vitro expansion of at least one stage. Further provided is a method for preventing and/or treating tumors by using the tumor-infiltrating lymphocyte.

Abstract: A script-to-movie generation method for a computing device includes: obtaining a movie script; generating a list of actions according to the movie script; generating stage performance based on each action in the list of actions; extracting observation information from the stage performance; using a camera agent trained with a reinforcement learning algorithm to select a camera based on the observation information, where the camera includes camera setting that defines a position of the camera with respect to a character for which the camera shoots; using the selected camera to capture a video of the stage performance; and outputting the video.

Abstract: A computer vision processor of a camera extracts attributes of persons or vehicles from hyperzooms generated from image frames. The hyperzooms represent traffic patterns. The extracting is performed using a feature extractor of an on-camera convolutional neural network (CNN) including an inverted residual structure. The attributes include at least colors of clothing of the persons or colors of the vehicles. Mobile semantic segmentation models of the CNN are generated using the hyperzooms and the attributes. Attribute analytics are generated by executing the mobile semantic segmentation models while obviating network usage by the camera. The attribute analytics are stored in a key-value database located on a memory card of the camera. A query is received from the server instance specifying one or more of the attributes. The attribute analytics are filtered using the one or more of the attributes to obtain a portion of the traffic patterns.

Abstract: The present disclosure provides an intracranial artery stenosis detection method and system. The present disclosure obtains artery stenosis detection results based on a first maximum intensity projection (MIP) image and a second MIP image obtained by preprocessing a medical image by adopting a detection model based on an adaptive triplet attention module and generates an auxiliary report and visualization results according to target category information in the artery stenosis detection results. Therefore, the problem that existing manual interpretation methods are easily affected by the subjective experience of doctors and are time-consuming and laborious can be solved, thus improving accuracy and efficiency of intracranial artery stenosis detection.

Abstract: A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.

Abstract: Embodiments of this application disclose a model training method performed by a computer device. The method includes: respectively performing, through m reference object key point detection models, object key point detection processing on a target training image to obtain key point detection results of predicted positions of key points respectively corresponding to a plurality of object key parts in a target training image; for each object key part whose associated predicted position of a key point corresponding to the object key part in the m key point detection results indicates that the object key part is a target object key part, determining a labeling position corresponding to the target object key part as a corresponding key point labeling result; and training a target object key point detection model using the target training image and the key point labeling results corresponding thereto.