Abstract: Implementations of the present specification provide a video retrieval method and apparatus. In the method, a video frame in a video to be matched is obtained; an image feature and a text feature are extracted from the video frame; the image feature and the text feature are fused based on a center variable used to represent a cluster center to obtain a fused feature, where the center variable is used to associate features of different modes of a same video; and video retrieval is performed in a video database based on the fused feature to determine a video in the video database that matches the video to be matched, where a plurality of videos and video features corresponding to the plurality of videos are stored in the video database.

Abstract: A terminal body includes a crimping part adapted to be crimped onto a cable, a mating part adapted to mate with a mating terminal, and a connecting part connected between the crimping part and the mating part. The crimping part includes a base connected to the connecting part and a pair of side wings extending from a pair of sides of the base beyond the connecting part. A stress relief notch is formed at a corner between each of the side wings and one of the connecting parts.

Abstract: Implementations of the present specification disclose methods, apparatuses, and devices for training an object detection system by using a gradient fine-tuning technique. In one aspect, the method includes: providing a training image as input to the object detection system; processing the training image by the object detection system; determining a gradient norm of each neural network layer based on object annotation data and the detection result corresponding to the training image; and updating, for each neural network layer, parameter values of the neural network layer based on an average of gradient norms and the gradient norm of the neural network layer.

Abstract: The present application relates to a floating connection conductor, a floating electric connector and a vehicle-mounted electronic device. The floating connection conductor comprises a welding part, a base-fixed part, a suspended floating part, a floating-fixed part and a contact part, wherein the base-fixed part is fixed to a base body; the floating-fixed part is fixed to a floating body, and the contact part is in conduction with and abuts against a plug connector; and the suspended floating part is located between the base-fixed part and the floating-fixed part and is bent to form a suspended section.

Abstract: Provided is a light transmitter, including: a substrate; a semiconductor refrigeration assembly arranged on the substrate; a light emitting assembly mounted on the semiconductor refrigeration assembly, so that the semiconductor refrigeration assembly cools the light emitting assembly, and the light emitting assembly is configured to generate a laser beam; a tube shell mounted on the substrate to package the semiconductor refrigeration assembly and the light emitting assembly; an optical fiber configured to output the laser beam generated by the light emitting assembly to an outside of the tube shell; and a transparent filling glue filled in a space between an inner wall of the tube shell and the substrate, and configured to guide the laser beam generated by the light emitting assembly to the optical fiber and transfer a heat generated by the light emitting assembly to the tube shell.

Abstract: A system includes a first hardware architecture and computational storage. The computational storage includes a second hardware architecture, and memory storing instructions that, when executed by the second hardware architecture, causes the system to perform a first set of operations. The first set of operations include receiving instructions from the first hardware architecture, performing one or more computations on data stored in the second hardware architecture, based on the received instructions, and transmitting a result to the first hardware architecture, based on the one or more performed computations.

Abstract: A buried thermistor includes a lower substrate, an upper substrate, and a number of thermistor stacks. Each thermistor stack includes two resistor subjects. Each resistor subject includes a base layer, a medium layer, a metal layer, a resistor layer, a nanometal layer, and a conductive layer. Applicable material of the resistor layer becomes more diverse by disposing the number of thermistor stacks, and the buried thermistor shows variable thermal sensitivity.

Abstract: The present invention discloses plating apparatuses and plating methods for metal deposition on a substrate with pattern structures. In an embodiment, a plating apparatus comprises a plating bath configured to accommodate electrolyte, a substrate holding module configured to hold a substrate, and at least one driving device configured to drive the substrate holding module together with the substrate to horizontally vibrate and/or vertically vibrate during the substrate being immersed into the electrolyte to be plated. The present invention can enhance mass transfer during the substrate being plated by vibrating the substrate holding module so as to raise plating rate and uniformly plating on the pattern structures.

Abstract: A short-range communication method and apparatus are applied to the field of short-range communication, particularly related to scenarios such as smart driving, smart manufacturing, and can establish time-frequency synchronization between nodes in an entire network. The method includes: A first node receives at least one piece of node information from at least one node; and the first node controls the first node to establish time-frequency synchronization with a second node that is in the at least one node and that belongs to a second synchronization area, where node information from the second node includes synchronization information of the second synchronization area, and the synchronization information indicates at least one of a node identifier, a synchronization status, or a synchronization direction between nodes in the second synchronization area. This solution may be further used to improve an autonomous driving or advanced driver assistance system ADAS capability.

Abstract: An apparatus to perform fast data access comprises a receiver, a processor, and a memory. The processor receives using the receiver a light signal from a light source. The light signal can be structured to generate a temporal code. The light source is an optical beacon that includes a Light-Emitting Diode (LED). The processor then decodes the light signal to generate a network address, and causes a display of a client device coupled to the apparatus to display information based on the network address. The network address can be a Uniform Resource Locator (URL) address and the information based on the network address includes a webpage associated with the URL. Other embodiments are described herein.

Abstract: A method and/or a communications apparatus controls a monitoring of a physical downlink control channel. The method includes: receiving, by a terminal device, a first message sent by a network device, where the first message includes M sleep time sets, and the sleep time is used to indicate a time length in which monitoring the PDCCH is stopped; receiving, by the terminal device, a second message sent by the network device, where the second message includes sleep time indication information, the sleep time indication information is used to indicate a target sleep time index, and the target sleep time index is an index in the M sleep time sets; and stopping, by the terminal device, monitoring the PDCCH in a target time period.

Abstract: This disclosure provides a beam management method and apparatus, to quickly implement beam alignment. The beam management method is applied to a first device and includes: obtaining a first reference signal sent by a second device based on a first set beam, and determining a first beam based thereon such that a signal quality of the first reference signal obtained by the first device based on the first beam is the best; and sending a second reference signal to the second device based on the first beam, where the second reference signal is used to determine a second beam of the second device, and a signal quality of the second reference signal obtained by the second device based on the second beam is the same as that of the first reference signal obtained by the first device based on the first beam.

Abstract: A method for analyzing a correlation between rail transit and direct current (DC) magnetic bias of a transformer includes the steps of, A: obtaining a current of a feed cable and a DC magnetic bias current: measuring the feed cable current in rail transit and the DC magnetic bias current of a transformer in a power grid within a certain period by a monitoring apparatus; B: calculating a characteristic quantity of the feed current within the measurement period based on the obtained current of the feed cable; C: calculating a characteristic quantity of the DC magnetic bias current within the measurement period based on the DC magnetic bias current; and D: calculating a support degree and a confidence coefficient based on the calculated characteristic quantity of the feed current and the calculated characteristic quantity of the DC magnetic bias current, and generating a correlation rule.

Abstract: Disclosed are a dual-optical-path spectrophotometer and a color measurement method thereof. The spectrophotometer includes an integrating sphere, a light source, and a sensor. A second shutter, a semi-reflecting and semi-transmitting device and lenses are arranged between the detection hole and the sensor, and a light guide device and a first shutter are arranged between a light guide hole formed in the integrating sphere and the semi-reflecting and semi-transmitting device. The color measurement method includes the following steps. A first shutter is closed, a second shutter is opened, light, reflected by the measuring opening, enters a sensor and the sensor measures a spectral reflected signal of the object surface. The first shutter is opened, the second shutter is closed, reflected light enters the sensor, and the sensor measures a spectral reflected signal of a light source. A final sampled signal is calculated.

Abstract: The present disclosure describes techniques for improving data processing. At least one transaction may be generated by a file system driver associated with a host. Each transaction comprises logical block addresses (LBAs) and information associated with at least one function. The at least one transaction may be journaled into a computational storage for offloading data processing from the host to the computational storage. The computational storage comprises a persistent memory and at least one field-programmable gate array (FPGA) core. The offloading of the data processing comprises offloading journal replay operations to the computational storage. The journal replay operations comprise applying the at least one function to at least one subset of the LBAs by the at least one FPGA core.

Abstract: The present disclosure describes techniques for guaranteeing online services based on predicting failures of storage devices. Statistical data may be extracted on a regular basis by each of a plurality of storage devices. Each of the plurality of storage devices may comprise a set of NAND dies. Each of the set of NAND dies may be configured to measure and track a set of metrics indicating characteristics of each NAND die. Prediction data indicating potential failures of the plurality of storage devices may be generated. The prediction data may be shared with a host on a periodic basis. A strategy of decommissioning an aged storage device and adding a new storage device based on the prediction data may be created by the host. The data migration to the new storage device may be implemented.

Abstract: This application discloses a neural network-based channel estimation method and a communication apparatus. The method may be implemented in the following steps: obtaining first location information of a first communication apparatus; processing the first location information by using a neural network, to obtain first channel information, where the first channel information is information about a radio channel between the first communication apparatus and a second communication apparatus; and performing communication based on the first channel information. The neural network is used to estimate channel information based on location information. This can avoid channel estimation based on a pilot signal, reduce overheads of the pilot signal, and improve efficiency and accuracy of estimating the channel information.

Abstract: A method and system for high-speed transient thermal simulation of an electronic device and belongs to the technical field of high-speed transient thermal simulation of electronic devices. According to acquired parameter data of the electronic device, a dynamic weak balance relationship among heat generation amount, internal energy increment and heat dissipation amount of the electronic device is calculated to obtain a functional relationship between operating temperature and time of the electronic device; a trough temperature value of a transient temperature curve of electronic device in weak balance state is obtained by limit solving algorithm; an initial temperature is set in a manner of loading a fixed-temperature heat source, and simulating calculation is performed for a first preset number of cycles to obtain an initial temperature field; and a high-speed transient temperature change of the electronic device is obtained by the operation of a second preset number of pulse stress cycles.

Abstract: A multi-type industrial robot control system may include: an automation controller; and a programmable logic controller (PLC)-robot bridge. The controller stores a general-purpose robot control function block generated using a unified design standard, calls a corresponding robot control function block for a robot and using robot state data and puts out corresponding first command data to the PLC-robot bridge. The PLC-robot bridge virtualizes robot interfaces of corresponding offline programming functions of different types of robot as a unified virtual interface, calls a corresponding offline programming function via the unified virtual interface according to the first command data, and generates second command data for output to a corresponding robot controller. The corresponding robot controller controls a corresponding robot. At the same time, robot state data from a corresponding type of robot is received via the unified virtual interface and the robot state data is fed back to the automation controller.

Abstract: Provided are an image detection method and apparatus, a device, a storage medium and a computer program product, the method includes: performing feature extraction on an image to be detected to obtain a feature map; performing target detection on the feature map through a target branch network to obtain target detection information of candidate anchor boxes at a pixel position of the feature map; performing keypoint detection on the feature map through a keypoint branch network to obtain positions of keypoints in the candidate anchor boxes at the pixel position of the feature map, and generating heatmap groups of the candidate anchor boxes according to the positions of the keypoints; and obtaining information about a target object in the image to be detected and positions of keypoints of the target object according to the target detection information and the heatmap groups.