Construction method and application of convolutional neural network based on attention mechanisms
A technology of convolutional neural network and construction method, which is applied to the construction of convolutional neural network based on attention mechanism, and the application field of convolutional neural network in image classification, which can solve the problems of low image recognition efficiency and recognition rate, and achieve Improve convergence efficiency and accuracy, improve network performance, and improve network performance
Active Publication Date: 2018-11-02
HUBEI UNIV OF TECH
View PDF8 Cites 20 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] Aiming at the problems of low image recognition efficiency and recognition rate of existing convolutional neural networks, the present invention proposes a convolutional neural network construction method based on attention mechanism and its application in image recognition
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0038] In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the implementation examples described here are only used to illustrate and explain the present invention, and are not intended to limit this invention.
[0039] An attention mechanism-based convolutional neural network construction method and application provided in an embodiment of the present invention include the following steps:
[0040] Step 1: Basic convolution operation:
[0041] Establish a basic convolution operation with 2 convolutional layers and 2 pooling layers. The convolutional layer and the pooling layer are spaced apart, and the network structure of this part is: convolutional layer (the size of the convolution kernel is 5*5 , the number of convolution kernels is 64, the convolution step is 1...
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
Login to View More Abstract
The invention discloses a construction method and application of convolutional neural network based on attention mechanisms. The convolutional neural network comprises a basic convolutional operationlayer for preprocessing images, an attention mechanism layer 1 for extracting obvious characteristics of superficial layers of images, an attention mechanism layer 2 for extracting obvious characteristics of deep layers of the images, an attention mechanism layer 3 for extracting obvious characteristics of the deepest layers of the images, two full connection layers for smoothing two dimensional output data in the attention mechanism layer 3 into one dimension output data, and a SoftMax classifier. According to the construction method and application of convolutional neural network based on the attention mechanisms disclosed by the invention, the attention mechanisms are combined with the convolutional neural network, so that extraction of effective information by the convolutional neuralnetwork is effectively promoted, and therefore, the network performance is improved, and the network convergence efficiency and precision are promoted.
Description
technical field [0001] The invention relates to the technical field of image processing, in particular to a method for constructing a convolutional neural network based on an attention mechanism and an application of the convolutional neural network in image classification. Background technique [0002] The visual attention mechanism is a brain signal processing mechanism unique to human vision. Human vision quickly scans the global image to obtain the target area that needs to be focused on, which is generally referred to as the focus of attention, and then invests more attention resources in this area to obtain more detailed information about the target that needs to be focused on. And suppress other useless information. [0003] One of the typical frameworks of modern computer vision, convolutional neural network, has been widely used in the field of image processing in recent years. One of its main applications is pattern recognition and image classification. At this s...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06N3/08G06N3/04G06K9/62
CPCG06N3/082G06N3/045G06F18/241
Inventor 王改华袁国亮吕朦刘文洲李涛
Owner HUBEI UNIV OF TECH