An aesthetic attribute evaluation method based on dense convolution network and multi-task network

A convolutional network and evaluation method technology, applied in the field of multi-task analysis and prediction, can solve the problems of high accuracy of aesthetic attribute prediction results, low accuracy of aesthetic attribute prediction results, and large amount of calculation, so as to achieve simple program writing and debugging , Strengthen the transmission of characteristic parameters, and have a wide range of applications

Inactive Publication Date: 2019-01-22
中共中央办公厅电子科技学院
View PDF8 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The disadvantage of the existing algorithms is that the amount of calculation is huge, and it takes a long time to train the data set to obtain the corresponding results. Secondly, for the predicted results, there are individual aesthetic attributes with high prediction results and other aesthetic attributes. The status quo of low prediction accuracy
As far as the method itself is concerned, there are still some deficiencies

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 aesthetic attribute evaluation method based on dense convolution network and multi-task network
  • An aesthetic attribute evaluation method based on dense convolution network and multi-task network
  • An aesthetic attribute evaluation method based on dense convolution network and multi-task network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0041] Dense Convolutional Neural Network:

[0042] In ResNet (a common classic convolutional network model, such as figure 1 ), the relationship between two adjacent layers can be expressed by the following formula:

[0043] x l =H l (X l-1 )+X l-1 (1)

[0044] where l represents the layer, X l Denotes the output of layer l, H l Represents a nonlinear transformation. So for ResNet, the output of layer l is the output of layer l-1 plus the nonlinear transformation of the output of layer l-1.

[0045] By changing the way information is transferred between layers, dense modules propose a new connection method. Any of these needs to be related to its subsequent layers. Its mathematical expression is as follows:

[0046] x l =H l ([X 0 ,X 1 ,...,X l-1 ]) (2)

[0047] where [X 0 ,X 1 ,...,X l-1 ] refers to the concate...

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 an aesthetic attribute evaluation method based on a dense convolution network and a multi-task network. Aesthetic features of images are extracted from dense convolution neuralnetwork model, and a part of image features are preserved in multi-dimensional matrix. The hierarchical multi-task network is used for regression analysis of known image attributes. After many times of training analysis, the prediction results and the data in the training data set reach a high degree of fitting, the final training model is saved, and the model is tested on the test data set to getthe regression results of the method. Because the data set used does not have a certain tendency, the aesthetic attribute prediction algorithm model has a certain universality. The method is implemented using Google's Tensorflow framework, and can be widely used in computer vision, image analysis and processing, digital photography and digital entertainment and other fields.

Description

technical field [0001] The invention belongs to the field of visual computing and computer vision, especially the field of image aesthetic evaluation, especially a multi-task analysis and prediction method. Background technique [0002] Early work on the evaluation of image aesthetic attributes mainly focused on the manual design of various image aesthetic features and the use of pattern recognition algorithms for aesthetic quality prediction. Another line of research attempts to directly fit the quality of image aesthetics through some hand-designed general image features. [0003] Recently, studies have shown promising performance from big data deep image features, and performance beyond traditional hand-designed features, and using image aesthetic attribute evaluation data from online professional photography communities. [0004] Aesthetic evaluation is a subjective vision task. Therefore, the quality evaluation of image aesthetics is ambiguous, and there are different...

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): G06K9/62G06K9/46G06N3/04
CPCG06V10/44G06N3/045G06F18/2193
Inventor 金鑫吴乐周兴晖赵耿张晓昆
Owner 中共中央办公厅电子科技学院
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