Target detection processing method, apparatus and device

The invention provides a target detection processing method, apparatus and device. The method comprises the steps of obtaining a first convolution result corresponding to a first convolution layer of a target detection model; performing low-bit quantization on the first convolution result to obtain...

Full description

Saved in:
Bibliographic Details
Main Authors ZHAO DONGYU, YUAN HAOQIANG, XIA LICHAO, DAI ZHENYUAN, CAO JIAN
Format Patent
LanguageChinese
English
Published 06.08.2021
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:The invention provides a target detection processing method, apparatus and device. The method comprises the steps of obtaining a first convolution result corresponding to a first convolution layer of a target detection model; performing low-bit quantization on the first convolution result to obtain a first quantization result; converting the first quantization result into an input working frame corresponding to a pulse neural network chip; sending the input working frame to the spiking neural network chip, so that the spiking neural network chip performs operation according to the input working frame to obtain an output working frame; determining a target detection result according to the output working frame. Thereby the memory occupation of the detection model can be effectively reduced, and the operation of the middle layer part of the target detection model by the spiking neural network chip can be realized, the neural network reasoning time can be effectively decreased, the data transmission between stor
Bibliography:Application Number: CN202110252763