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

Method and device for performing convolution calculation on image by using convolutional neural network

A convolutional neural network and convolution operation technology, applied in the field of convolutional neural network, can solve the problems of low efficiency, high power consumption, slow convolutional neural network speed, etc., to reduce power consumption, improve efficiency, and reduce the amount of calculation Effect

Inactive Publication Date: 2020-07-14
AXERA TECH (BEIJING) CO LTD
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

From the beginning to the end, the entire model needs to calculate all the regions of the input image. When the input image is large, to obtain sufficient accuracy, the overall computing power of the model will be particularly large, resulting in a slow convolutional neural network. , lower efficiency, higher power consumption

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
  • Method and device for performing convolution calculation on image by using convolutional neural network
  • Method and device for performing convolution calculation on image by using convolutional neural network
  • Method and device for performing convolution calculation on image by using convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] First, refer to figure 1 An example electronic system 100 for implementing the method and apparatus for performing convolution calculation on an image using a convolutional neural network according to an embodiment of the present invention will be described.

[0031] Such as figure 1 A schematic structural diagram of an electronic system is shown, the electronic system 100 includes one or more processing devices 102, one or more storage devices 104, input devices 106, output devices 108 and one or more image acquisition devices 110, these components The interconnections are via bus system 112 and / or other forms of connection mechanisms (not shown). It should be noted that figure 1 The components and structures of the electronic system 100 shown are exemplary rather than limiting, and the electronic system may also have other components and structures as required.

[0032]The processing device 102 may be a gateway, or an intelligent terminal, or a device including a c...

Embodiment 2

[0039] This embodiment provides a method for performing convolution calculation on an image using a convolutional neural network. In the method, the image is first input into the convolutional neural network; the first convolutional layer of the convolutional neural network outputs the first feature map, and the first convolutional neural network outputs the first feature map. A feature map includes a first control channel, and the first convolutional layer is an intermediate convolutional layer of a convolutional neural network.

[0040] The convolutional neural network in this embodiment is mainly used to classify images, target detection or target recognition. The calculation of the convolutional neural network can extract the characteristics of people or objects in the image, and identify the people or objects in the image. Or count the people or objects in the image, etc.

[0041] A convolutional neural network usually consists of one or more convolutional layers and a fu...

Embodiment 3

[0056] This embodiment provides another method for performing convolution calculations on images using a convolutional neural network, which is implemented on the basis of the above-mentioned embodiments; this embodiment focuses on the redundant process of determining the first feature map according to the first control region. The specific process of the remaining area. Such as image 3 The flow chart of another method for performing convolution calculation on an image using a convolutional neural network is shown, including the following steps:

[0057] Step S302, taking the area in the first control channel whose eigenvalue is smaller than the preset threshold value as the first control area, the above-mentioned first control channel is a channel in the first feature map, and the eigenvalue of any position of the above-mentioned first control channel is used The probability that the corresponding position in the first feature map belongs to the target feature.

[0058] Th...

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 provides a method and a device for performing convolution calculation on an image by using a convolutional neural network. The image is input into a convolutional neural network; and thefirst convolutional layer outputs a first feature map, the first feature map comprises a first control channel, and the first convolutional layer is a middle convolutional layer of the convolutionalneural network. The method comprises: taking an area with a characteristic value smaller than a preset threshold value in a first control channel as a first control area, wherein the characteristic value of any position of the first control channel is used for representing the probability that the corresponding position of a first characteristic graph belongs to target characteristics; determininga redundant area of the first feature map according to the first control area; and enabling other regions except the redundant region in the first feature map to enter the next convolution layer of the first convolution layer for convolution operation so as to obtain a second feature map. In the mode, the redundant area does not perform the next layer of convolution operation, invalid calculationcan be prevented, the calculation amount of the neural network is reduced, the convolution calculation efficiency is improved, and the power consumption of convolution calculation is reduced.

Description

technical field [0001] The present invention relates to the technical field of convolutional neural networks, in particular to a method and device for performing convolution calculations on images using convolutional neural networks. Background technique [0002] In the related art, the convolutional neural network algorithm includes a one-step method and a two-step method. Among them, the two-step method requires two models for calculation. First, the first model is used to calculate the rough detection network to obtain the effective area, such as the probability heat map of the appearance of the target (such as a face); and then the output of the first model The image is matted on the basis of the image, and the matted image is used as the input image of the second model, and the second model performs the calculation of the fine detection network or attribute network, so as to obtain the required accurate detection or attribute results. The two-step convolutional neural ...

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): G06N3/04G06N3/063G06N3/08G06T1/20
CPCG06N3/063G06N3/084G06T1/20G06N3/045
Inventor 梁喆周舒畅曹宇辉
Owner AXERA TECH (BEIJING) CO LTD
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