Unlock instant, AI-driven research and patent intelligence for your innovation.

A neural network-based target detection method and device

A target detection and neural network technology, applied in the field of deep learning, can solve problems such as consuming large computing power and bandwidth, and achieve the effect of reducing computing power consumption

Active Publication Date: 2022-04-19
SHENZHEN MICROBT ELECTRONICS TECH CO LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Whether it is the calculation of the confidence level or the calculation of the coordinate position, it takes a lot of computing power and bandwidth

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
  • A neural network-based target detection method and device
  • A neural network-based target detection method and device
  • A neural network-based target detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] In order to make the purpose, technical means and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings.

[0062] This application uses the causal relationship between the confidence degree and the coordinate position, calculates the confidence degree in advance, and uses the reverse operation of the first operation for processing the organizational form of the feature data in the confidence degree calculation process to determine the feature point. coordinate location. This application can greatly reduce the demand for coordinate position calculation, thereby reducing the demand for computing power and bandwidth.

[0063] see Figure 4 as shown, Figure 4 It is a schematic flow chart of the object detection method based on neural network in this application. The method includes,

[0064] Step 401, performing a first operation on the feature values ​​of the feature...

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 present application discloses a target detection method based on a neural network. The neural network performs target detection on the input data, and performs a first operation on the feature values ​​of the feature points in the output result of the convolution operation of the neural network, wherein the first operation It is used to process the organizational form of the feature value, and calculate the confidence degree of the feature point based on the result of the first operation; according to the first position information of the feature point in the result of the first operation, through the reverse operation of the first operation , to obtain the second position information of the feature point in the output result of the convolution operation, and output the confidence degree of the feature point and the second position information of the feature point to obtain the target detection result. This application reduces computing power consumption in the detection process.

Description

technical field [0001] The present invention relates to the field of deep learning, in particular, to a neural network-based target detection method and device. Background technique [0002] In machine learning and deep learning, object detection based on neural networks has a wide range of applications. Although the target detection effect based on neural network is good, it relies on a large amount of calculation, which leads to the consumption of computing power and bandwidth on the hardware. [0003] Take the one-stage multi-target detection method (SSD, Single Shot Multi-Box Detector) in the deep learning detection network as an example. see figure 1 as shown, figure 1 It is a schematic diagram of SSD neural network structure. The neural network structure of SSD includes basic network and pyramidal network. The basic network is the first 4 layers of the Visual Geometry Group (VGG) such as VGG-16. Pyramid networks are simple convolutional networks with progressivel...

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 Patents(China)
IPC IPC(8): G06V10/25G06V10/82G06N3/04G06F40/279G10L25/30
CPCG06F40/279G10L25/30G06N3/045
Inventor 张宁杨作兴房汝明向志宏
Owner SHENZHEN MICROBT ELECTRONICS TECH CO LTD