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

Semantic inference-based glancing path prediction method

A prediction method and path technology, applied in the direction of neural learning methods, instruments, biological neural network models, etc., can solve the problems of ambiguity in physiological mechanism, modeling unable to adapt to the scanning path of different groups of people, etc., and achieve strong semantic representation ability, strong The effect of semantic abstraction ability

Pending Publication Date: 2021-08-27
NORTHWESTERN POLYTECHNICAL UNIV
View PDF1 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Third, there is ambiguity in the physiological mechanism of the saccade path, and the existing models rely on a single physiological modeling and cannot adapt to the saccade path of different groups of people

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
  • Semantic inference-based glancing path prediction method
  • Semantic inference-based glancing path prediction method
  • Semantic inference-based glancing path prediction method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] Step 1, build a test image library

[0053] Collect images and capture the corresponding glance path of the image, and uniformly transform the size of all images to h×w size, where h represents the height of the image, w represents the width of the image, and each image corresponds to the size conversion coefficient r x 、r y . Calculate the pixel corresponding to the fixation point of the glance path after transformation, let (q 1 ,q 2 ,...,q n ) represents the gaze point coordinate sequence of the glance path.

[0054] Step 2, build a semantic extractor to extract the semantic features of gaze points

[0055] The semantic extractor realizes the mapping from image block pixels to high-level semantic information. CNN is suitable for processing image data, so the commonly used CNN model VGG-16 is selected as the semantic extractor, and the model parameters are trained in the classification task. The compression factor of the selected image size h / h'=8, then select t...

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 relates to a semantic inference-based glancing path prediction method, and belongs to the field of image glancing path prediction. An image glancing path training set is constructed, and the image is mapped to a semantic space by using a trained CNN to obtain a semantic vector corresponding to the fixation point. A glancing path prediction model of an encoder-decoder framework is constructed, global information of an encoder encoding image outputs an encoding vector, the encoding vector is used as an initial value of a decoder, the decoder learns a fixation point semantic inference relation, an Euclidean distance from a prediction fixation point semantic vector to a true value fixation point semantic vector is used as a loss function, and the encoder-decoder network is optimized such that the loss function is minimized. And an image is input in the optimized network for testing to obtain a glancing path.

Description

technical field [0001] The present invention relates to the field of image glance path prediction, in particular to a glance path prediction method based on semantic inference, that is, to map glance points to semantic space, and then establish the semantic jump relationship of glance points through supervised learning, so as to realize the prediction of glance path. Background technique [0002] The human eye receives a large amount of visual data that is far beyond the processing of the human brain at all times. The human visual system can find important areas from complex physical scenes, which enables people to use less computing resources in a large number of visual data. Get useful information quickly from the data. Therefore, the study of the human visual system is of great significance to how to quickly extract useful information from a large amount of visual data. At present, the existing research is divided into two aspects: one is visual salience, which means the...

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): G06K9/34G06K9/62G06N3/04G06N3/08
CPCG06N3/049G06N3/08G06V10/267G06N3/045G06F18/213G06F18/214
Inventor 夏辰钟文琦韩军伟郭雷
Owner NORTHWESTERN POLYTECHNICAL UNIV
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