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

Training methods for convolutional neural network models for image processing

A convolutional neural network and image processing technology, applied in the field of image processing, can solve problems such as short computing time and tight computing resources

Active Publication Date: 2022-07-08
SHENZHEN MICROBT ELECTRONICS TECH CO LTD
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, when image processing is performed on the smart device side, computing resources are usually tight, and the expected computing time is usually short

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
  • Training methods for convolutional neural network models for image processing
  • Training methods for convolutional neural network models for image processing
  • Training methods for convolutional neural network models for image processing

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] Various exemplary embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. It should be noted that the relative arrangement of the components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0023] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application or uses in any way. That is, the structures and methods herein are shown by way of example to illustrate various embodiments of the structures and methods in the present disclosure. Those skilled in the art will appreciate, however, that they are merely illustrative, and not exhaustive, of the ways in which the disclosure may be practiced. Furthermore, the figures are not necessarily to scale and some features may be exaggerate...

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 disclosure relates to a training method of a convolutional neural network model for image processing. A method for training a convolutional neural network model for image processing, comprising: using a training set image to train a convolutional neural network model to be trained, the convolutional neural network model including a convolution kernel including a plurality of channels, used for Perform convolution processing on image features of multiple channels; and verify the trained convolutional neural network model: cluster the channel vectors of image features based on test set images to obtain reference channel vectors; Each convolution kernel that performs convolution processing calculates the dot product result of each channel vector and each reference channel vector to obtain a channel vector dot product table; and stores the reference channel vector and the channel vector dot product corresponding to the convolution kernel. Tables are the parameters of the trained convolutional neural network model, and the channel vector is a vector composed of values ​​at the same coordinates in different channels in the image features or convolution kernels.

Description

technical field [0001] The present disclosure relates to image processing on the smart device side. [0002] Specifically, it relates to a training method for a convolutional neural network model used for image processing on a smart device side. The image processing method using the convolutional neural network model trained in this way, on which the computer storage medium of the above method is stored, realizes The image processing apparatus of the above method and the smart device including the image processing apparatus. Background technique [0003] At present, in smart devices (such as mobile phones, tablet computers, smart cameras, smart gates, etc.), there is a wide range of needs for image processing technology. For example, in a smart camera, image processing is required to realize functions such as face recognition and beauty. [0004] Existing image processing methods use a convolutional neural network model, which requires a lot of computation during image pro...

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): G06V40/16G06V10/774G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/214
Inventor 艾国杨作兴房汝明向志宏
Owner SHENZHEN MICROBT ELECTRONICS TECH CO LTD