An image recognition method based on a self-adaptive full convolution attention network

A technology of image recognition and attention, applied in character and pattern recognition, instruments, computer parts, etc., can solve problems such as model failure and expensive

Inactive Publication Date: 2019-05-10
TAIYUAN UNIV OF TECH
View PDF1 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Most of the traditional methods use human to locate the region of interest, which has the following disadvantages: 1) Accurate annotations are very expensive; 2) If the annotations are not accurate, the model will fail; 3) Finally But most importantly, don't know how to manually define the optimal recognition part

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
  • An image recognition method based on a self-adaptive full convolution attention network
  • An image recognition method based on a self-adaptive full convolution attention network
  • An image recognition method based on a self-adaptive full convolution attention network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] Attached below Figure 1-3 , the image recognition method based on the self-adaptive full convolution attention network of the present invention, comprises the following steps as follows:

[0054] 1. Training the neural network

[0055] 1. Activation function

[0056] When selecting the activation function, considering that the sigma function and the tanh function will have a gradient disappearance problem, the ReLU function is highly efficient in the area where the input is positive. In practice, it is more reasonable to converge faster than sigma / tanh, but When it is less than 0, the gradient is 0. At this time, we introduced the PReLU function, which can effectively solve the above problems. The PReLU function is defined as follows: f(x)=max(ax,x); a is a coefficient, and the function value represents the maximum value of input x and ax;

[0057] 2. Data preprocessing - feature compression

[0058] Too many features of the picture will cause great troubles to the...

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 an image recognition method based on a self-adaptive full convolution attention network. The invention discloses a self-adaptive full convolution attention network-based imagerecognition method, belongs to the technical field of image recognition methods, and aims to provide a method for improving the image recognition speed and accuracy by using the self-adaptive full convolution attention network and further applying a trained model. According to the technical scheme, the image recognition method based on the self-adaptive full convolution attention network comprisesthe following steps of 1, training a neural network: a, selecting an activation function; b, a data preprocessing step; b, adaptive learning rate; d, defining a network structure: (1) training a fullconvolution attention network; (2) training an adaptive image; Secondly, the model trained in the first step acts on a target image set, the data obtained in the second step is compared with the target image set, and therefore images are recognized and classified; The method is applicable to the field of image recognition.

Description

technical field [0001] The invention belongs to the technical field of image recognition methods, in particular to an image recognition method based on an adaptive full convolution attention network. Background technique [0002] Computers rival human experts in visual understanding. Compared with general image recognition, fast and accurate image recognition is more challenging, because the image features we pay attention to are not on the entire image, but in a certain area of ​​the image. Therefore, this requires us to simultaneously locate and describe the ROI in detail. Most of the traditional methods use human to locate the region of interest, which has the following disadvantages: 1) Accurate annotations are very expensive; 2) If the annotations are not accurate, the model will fail; 3) Finally But most importantly, don't know how to manually define the optimal part of recognition. Contents of the invention [0003] The present invention overcomes the deficiencie...

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): G06K9/62G06K9/46
Inventor 李灯熬赵菊敏白小红巩建平
Owner TAIYUAN UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap