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

A Line of Sight Estimation Method Based on Deep Regression Network

A technology of line of sight estimation and line of sight direction. It is applied in the field of safe driving, region of interest detection, and computer vision. It can solve problems such as difficult to accurately describe the relationship between human eye features and line of sight mapping.

Inactive Publication Date: 2019-02-15
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The main advantage of the regression-based line of sight estimation method is that it is simple and easy to implement under the premise of accurate positioning of the human eye area, but the disadvantage is that it is difficult for the existing regression methods to describe the mapping relationship between human eye features and line of sight very accurately.

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
  • A Line of Sight Estimation Method Based on Deep Regression Network
  • A Line of Sight Estimation Method Based on Deep Regression Network
  • A Line of Sight Estimation Method Based on Deep Regression Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0057] According to the method of the present invention, at first utilize Matlab or C language to write the training model of depth regression network; Then import the training sample that gathers and train the depth regression network parameter; Then extract gradient direction histogram feature to the image that gathers, as source data Input to the trained deep regression network for processing; get the estimated line of sight direction. The method of the present invention can be used in the problem of eye sight estimation in natural scenes.

[0058] A line-of-sight estimation method based on deep regression network, which comprises the following steps:

[0059] Step 1: Collect N eye images containing different sight lines (see figure 1 ), and record the corresponding line-of-sight direction when capturing each image the y n The first dimension of represents the horizontal direction, the second dimension represents the vertical direction, and the subscript n represents 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

This patent proposes a line-of-sight estimation method based on a deep regression network, which belongs to the field of computer vision and machine learning. The main idea of ​​this method is to establish the mapping relationship between the input image features and the line of sight through a deep regression network. First, extract the gradient direction histogram feature of the eye image area; then, establish a 5-layer depth regression model to fit the mapping relationship between the input image features and the output line of sight direction; after that, use the gradient descent method to optimize the depth regression model. parameters; finally, for the eye image to be estimated, use the learned depth model to estimate the line of sight direction.

Description

technical field [0001] The invention belongs to the technical field of computer vision, relates to a deep learning method, mainly solves visual mapping problems such as line of sight estimation and line of sight tracking, and can be applied to fields such as safe driving of automobiles and region of interest detection. Background technique [0002] In computer vision, gaze estimation refers to locating the human eye area based on the input facial image and automatically estimating the gaze direction according to the position of the eyeball. Existing gaze estimation methods include two categories: (1) geometry-based methods and (2) regression-based methods. See references for details: Takahiro Ishikawa, Simon Baker, Iain Matthews, and Takeo Kanade, Passive Driver Gaze Tracking with Active Appearance Models, tech.report CMU-RI-TR-04-08, 2004. [0003] The gaze estimation method based on geometry mainly calculates the gaze direction by locating the center of the pupil, the fea...

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): G06N3/08G06K9/46
CPCG06N3/084G06V10/50
Inventor 潘力立
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
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