Abstract: This application discloses a unit structure of a silicon photomultiplier tube, including a first conductive type heavily doped first electrode region located on a first side of the shallow trench isolation, a second conductive type heavily doped second electrode region located on a second side, and a quenching resistor located on a top surface of the shallow trench isolation. A photosensitive layer is formed in the silicon substrate at bottoms of the first electrode region, the shallow trench isolation and the second electrode region. The first electrode region, the photosensitive layer and the second electrode region form a Geiger mode avalanche photodiode. A first end of the quenching resistor is connected to the first electrode region through a first metal interconnect structure. A second end of the quenching resistor is connected to a first electrode. The second electrode region is connected to a second electrode.

Abstract: A method for processing overhead of memory access includes: applying for a memory configured to perform value padding on at least one convolution operation in a deep learning model; determining input data of the deep learning model; performing deep learning processing on the input data by using the deep learning model; and releasing the memory after performing the deep learning processing.

Abstract: An array substrate, a testing method and manufacturing method thereof, and a displaying device. According to the disclosure, a plurality of test capacitors connected in parallel, a first connecting wire, a second connecting wire, a first test terminal and a second test terminal are arranged in an array substrate located in a test area, each test capacitor comprises a first electrode, a first insulating layer and a second electrode which are stacked on a base, and each test capacitor has the same film structure as a storage capacitor arranged in an active area. By making the first electrode of each test capacitor in the test area connected to the first test terminal to through the first connecting wire, and the second electrode of each test capacitor connected to the second test terminal through the second connecting wire, the plurality of test capacitors are connected in parallel.

Abstract: The present disclosure provides a method and device for determining an optimal operation path. The method can include determining a model operation topology graph according to a neural network model, the model operation topology graph including a plurality of operation nodes, and selecting a plurality of operators with a same operation type as the operation node for each operation node in the model operation topology graph, and creating a corresponding operator node for each operator. Further, the method can include, according to a connection relationship between the operation nodes in the model operation topology graph, constructing an operator topology graph based on the created operator nodes, determining an operation index of each operator node, and weighting the operator topology graph based on the operation index of each operator node to obtain a weighted operator topology graph, and selecting an optimal operation path in the weighted operator topology graph.

Abstract: A method for processing overhead of memory access includes: applying for a memory configured to perform value padding on at least one convolution operation in a deep learning model; determining input data of the deep learning model; performing deep learning processing on the input data by using the deep learning model; and releasing the memory after performing the deep learning processing.

Abstract: An encrypted model file is acquired by encrypting at least a part of model information in an original model file. The original model file describes a target neural network model. A model program code is generated according to the encrypted model file. The model program code describes the target neural network model. An installation package for installing an application (APP) is sent to a User Equipment (UE) based on a request sent by the UE. The installation package includes the model program code.

Abstract: A method for remote assistance is provided. The method includes: receiving a remote assistance file sent by a controlling terminal, the remote assistance file containing a script file generated by the controlling terminal based on user operation; determining whether the script file is compatible to the controlled terminal; and if the script file is compatible to the controlled terminal, executing the script file to perform the user operation corresponding to the script file. Thereby, by transmitting a script file generated based on user operation between the controlled terminal and the controlling terminal, remote assistance can be performed more conveniently and user experience can be improved.

Abstract: A method for remote assistance is provided. The method includes: receiving a remote assistance file sent by a controlling terminal, the remote assistance file containing a script file generated by the controlling terminal based on user operation; determining whether the script file is compatible to the controlled terminal; and if the script file is compatible to the controlled terminal, executing the script file to perform the user operation corresponding to the script file. Thereby, by transmitting a script file generated based on user operation between the controlled terminal and the controlling terminal, remote assistance can be performed more conveniently and user experience can be improved.