Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Target detection method, computer equipment and readable storage medium

A target detection and computer program technology, applied in the field of target detection, can solve the problems of diverse value ranges, difficult to count quantitative values, uncontrollable lighting conditions, etc., to reduce quantization loss, reduce trainable parameters, and reduce quantization. range effect

Inactive Publication Date: 2020-08-11
浙江芯劢微电子股份有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing technology has the following defects: first, in practical applications, the scene is complex and changeable, the lighting conditions are uncontrollable, and the range of activation values ​​is varied, so it is difficult to calculate a quantitative value that meets the conditions of each scene; second , use asymmetric quantization to improve the expression accuracy of quantized values, but the calculation process after quantization becomes complicated

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Target detection method, computer equipment and readable storage medium
  • Target detection method, computer equipment and readable storage medium
  • Target detection method, computer equipment and readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] Such as figure 1 As shown, the present embodiment provides a target detection method, specifically, a target detection method for face recognition, including the following steps:

[0051] Collecting training samples, the collected training samples include face training samples of various postures, angles, and different illuminations; preprocessing the collected training samples. The targets in the training samples include sub-targets. Specifically, in this embodiment, the sub-targets refer to the left eye, right eye, nose, left mouth corner, and right mouth corner of the human face. Preprocessing the collected training samples includes:

[0052] Map the corresponding position of the sub-target to the specified position, that is, map to the following standard positions according to the positions of the left eye, right eye, nose, left mouth corner, and right mouth corner:

[0053] Left eye: [30.29459953, 51.69630051];

[0054] Right eye: [65.53179932, 51.50139999];

[...

Embodiment 2

[0095] This embodiment provides a computer device, including a memory and a processor, where a computer program is stored in the memory, and the processor implements the method in any of the above embodiments when executing the computer program. Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer programs. Accordingly, the computer program can be stored in a non-volatile computer-readable storage medium, and when the computer program is executed, the method of any one of the above-mentioned embodiments can be realized. Wherein, any references to memory, storage, database or other media used in the various embodiments provided in the present application may include non-volatile and / or volatile memory. Nonvolatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a target detection method, computer equipment and a readable storage medium, and relates to the technical field of target detection. According to the technical scheme providedby the invention, when a training sample is used for training a target detection model, the input values of the convolution weight and the activation function are respectively quantified, the loss function is constructed according to the input values, and the quantified convolution weight and the quantified input value are adjusted and iterated. According to the method provided by the invention, the weight value is adjusted in the training process, the precision can be ensured, the requirements under various scene conditions are met, and the calculation process is simplified.

Description

technical field [0001] The invention relates to the technical field of target detection, in particular to a target detection method, computer equipment and a readable storage medium. Background technique [0002] At present, deep learning technology is developing rapidly, and key breakthroughs have been made in the fields of face detection and recognition, reaching or even surpassing the human level. Various application functions based on deep learning emerge in endlessly and sink into various industries. The model obtained based on deep learning training has good results, but it also has the disadvantage of a large amount of calculation, which limits the application on embedded devices with low power consumption and low computing power. [0003] In the existing technology, by quantizing the weight and activation value to a preset size, the hardware fixed-point operation speed is faster than the floating-point operation speed to solve the problem of large amount of calculat...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V40/161G06N3/045G06F18/214
Inventor 王增冠
Owner 浙江芯劢微电子股份有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com