Check patentability & draft patents in minutes with Patsnap Eureka AI!

A fine -grained exercise behavior recognition method based on object -level trajectory

A recognition method and fine-grained technology, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve the problem of not being able to capture timing information well, and achieve the effect of improving accuracy

Active Publication Date: 2022-08-05
NANJING UNIV OF POSTS & TELECOMM
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] (3) Behavior recognition based on three-dimensional convolutional network (C3D): the traditional 2D (the dimension of the convolution kernel is two-dimensional) convolution cannot capture the information on the time series well, and the 3D convolution passes the three-dimensional convolution kernel at the same time capture information in space and time

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 fine -grained exercise behavior recognition method based on object -level trajectory
  • A fine -grained exercise behavior recognition method based on object -level trajectory
  • A fine -grained exercise behavior recognition method based on object -level trajectory

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0080] Step 1: For each image in the training video of the football stop event, use YOLOV3 to detect the player and the football, and obtain their position information;

[0081] Step 2: Perform secondary interpolation on the detection results to achieve two goals: to complete the missing frames without changing the normal trajectory of the ball; to unify the effective frame length of all events to 45 frames;

[0082] Step 3: Extract the object-level trajectory features and the instantaneous movement features of the ball for each event, as the basis for judging the success or failure of the ball stop event;

[0083] Step 4: Divide the stop event set into training set and test set, take the 45-dimensional feature vector of step 3 as input, and the event label - stop success or failure as output, and perform LSTM training.

[0084] The object-level trajectory-based behavior recognition method proposed by the present invention specifically describes the interaction between people ...

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 fine-grained behavior recognition method based on object-level trajectories, which includes the following processes: S1, detecting the positions of players and footballs from each image of a football stop event video; S2, intercepting a fixed number of frames from the video As a valid frame; S3, extract the object-level trajectory feature and the instantaneous movement feature of the ball from each valid frame; S4, input the object-level trajectory feature and the instantaneous movement feature of the ball of all valid frames into the classifier to judge the stop event. whether succeed. The present invention can effectively identify fine-grained behaviors by introducing object-level trajectories to specifically describe the interaction between people and objects.

Description

technical field [0001] The invention relates to the technical field of fine-grained behavior recognition, in particular to a fine-grained behavior recognition method based on object-level trajectory. Background technique [0002] With the development of scientific and technological information technology and the continuous improvement of people's needs, computer vision and image processing technology have been widely used in many fields. Visual detectors are all over our life, they record massive video data, how to convert these videos into effective data and be used by people has become a hot research direction in the current video field. Due to the characteristics of fine-grained behavior, there are a lot of repetitions in the background between different categories, and the difference is not obvious. The fine-grained behavior recognition technology based on object-level trajectories will play an important role in the field of behavior recognition. [0003] Scholars at ho...

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): G06V40/20G06V20/40G06V10/764G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V40/20G06V20/44G06V20/46G06V20/41G06N3/044G06N3/045G06F18/24
Inventor 熊健王姮冰路丽果
Owner NANJING UNIV OF POSTS & TELECOMM
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com