Contrast-based weak supervision time sequence action positioning evaluation method and system

An evaluation method and contrast technology, which can be used in instruments, character and pattern recognition, computer parts, etc., and can solve problems such as unsatisfactory positioning effects.

Active Publication Date: 2019-09-13
XI AN JIAOTONG UNIV
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method and system for positioning and evaluating weakly-supervised time-series action positioning based on contrast, so as to solve the above-mentioned technical problem that the positioning effect under weak-supervised conditions is not ideal

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
  • Contrast-based weak supervision time sequence action positioning evaluation method and system
  • Contrast-based weak supervision time sequence action positioning evaluation method and system
  • Contrast-based weak supervision time sequence action positioning evaluation method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments.

[0055] see figure 1 and figure 2 , a contrast-based weakly supervised time-series action location evaluation method according to an embodiment of the present invention, specifically comprising the following steps:

[0056] Step 1: Action Classification Prediction at Known Segment Level and segment-level attention score prediction In the case of , the maximum pooling is used to obtain the local maximum value and its temporal position of the target category action classification prediction. Similarly, local minima and corresponding timing positions are also obtained.

[0057] The specific steps in step 1 include:

[0058] (1) For the prediction of action classification at the known segment level Leveraging Segment-Level Attention Score Prediction For preprocessing, if the attention score for the t position is less than a c...

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 discloses a contrast-based weak supervision time sequence action positioning evaluation method and system, and belongs to the field of computer vision and mode recognition. The contrast-based weak supervision time sequence action positioning evaluation method comprises the steps: carrying out the modeling on the contrast information in a video; then utilizing the contrast to obtain afragment-level edge degree measurement; and finally, scoring any candidate video segment by using the obtained edge degree and a known segment level action classification result. According to the contrast-based weak supervision time sequence action positioning evaluation method, three aspects of starting, ending and content are comprehensively considered, and finally, an evaluation result with relatively high reliability is given; and by means of the evaluation result, the time sequence action positioning accuracy can be greatly improved. According to the contrast-based weak supervision timesequence action positioning evaluation method, the action positioning effect of any video time period can be comprehensively evaluated under the weak supervision condition, and only video-level category labeling is needed, and data labeling does not depend on complex action boundaries, and the labeling burden of the data can be greatly reduced.

Description

technical field [0001] The invention belongs to the technical field of computer vision and pattern recognition, and in particular relates to a contrast-based weakly supervised temporal sequence action location evaluation method and system. Background technique [0002] With the popularity of shooting equipment and networks, the volume of video data is growing rapidly, and society's demand for intelligent video analysis technology is becoming more and more urgent. Among them, the issue of timing action positioning has always been a hot research issue, and has important applications in intelligent monitoring, action search, and automatic video summarization. How to find the target action contained in the video and give the accurate time boundary of the start and end of the action is still a difficult problem; one of the difficulties of this problem is that the collection of training data is difficult, which is different from the problem of action recognition, which can be simp...

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): G06K9/00G06K9/62
CPCG06V40/20G06V20/42G06F18/2155G06F18/24
Inventor 王乐刘子熠郑南宁
Owner XI AN JIAOTONG UNIV
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