Check patentability & draft patents in minutes with Patsnap Eureka AI!

Sampling box generation method, training method and neural network

A neural network and training method technology, applied in the generation method of sampling frame, training method and neural network field, can solve the problems of low accuracy, linear convolution accuracy can not meet actual needs, etc. Avoid gradient disappearance and improve the effect of information

Pending Publication Date: 2020-10-02
玖壹叁陆零医学科技南京有限公司
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, the sampling frame can be generated by linear convolution. Linear convolution is relatively easy to implement, but its accuracy is not very high.
With the improvement of practical application standards, the accuracy of linear convolution has gradually failed to meet actual needs.

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
  • Sampling box generation method, training method and neural network
  • Sampling box generation method, training method and neural network
  • Sampling box generation method, training method and neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.

[0046] It should be noted that like numerals and letters denote similar items in the following figures, therefore, once an item is defined in one figure, it does not require further definition and explanation in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second" and the like are only used to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0047] see figure 1 , the embodiment of the present application provides a neural network 100, the neural network 100 can be deployed on an electronic device such as a terminal or a server, and the neural network 100 can include: an input layer, a feature extraction layer, and an output layer.

[0048] Wherein, the input layer 110 is used to obtain an image t...

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

Embodiments of the invention provide a sampling box generation method, a training method and a neural network. The method comprises: obtaining a to-be-processed image containing an object; performingfeature extraction on the to-be-processed image by using a preset neural network to obtain a plurality of feature images with different scales; generating a sampling frame of each object in each feature image by using a neural network, and restoring the sampling frames to the to-be-processed image according to corresponding scales; and carrying out de-duplication on each sampling frame in the to-be-processed image, so that a sampling frame framing the object can be reserved for each object. A plurality of feature images with different scales are generated in the feature extraction process, anda sampling box is generated for the same object in each feature image. And when the sampling frames are returned to the original image, each object corresponds to a plurality of sampling frames. At the moment, the sampling frame capable of framing the object can be reserved by removing the weight of the sampling frame, so that the sampling frame of the object can be generated more accurately.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, in particular, to a method for generating a sampling frame, a training method and a neural network. Background technique [0002] In some application scenarios of image recognition, such as in the scene of object tracking, it is necessary to generate a sampling frame in the image to select the object frame to achieve continuous tracking of the object, such as in DM (Data Matrix, two-dimensional code) In the scene of code recognition, it is necessary to generate a sampling frame to select the area where the DM code in the image is located, so as to mark the location of the DM code, which is convenient for users to identify and refer to. [0003] At present, the sampling frame can be generated by linear convolution. Linear convolution is relatively easy to implement, but its accuracy is not very high. With the improvement of practical application standards, the accuracy o...

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
IPC IPC(8): G06K9/46G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V10/40G06N3/045G06F18/253
Inventor 常江龙毛立王志鹏孙明建
Owner 玖壹叁陆零医学科技南京有限公司
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com