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

A human motion prediction method based on attention mechanism

A technology of human action and prediction method, applied in artificial life, biological neural network models, instruments, etc., can solve problems such as stability and accuracy, and achieve the effect of avoiding information redundancy

Active Publication Date: 2018-12-25
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF5 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0012] The purpose of the present invention is to propose a new human body action prediction method in order to solve the problem of stability and accuracy when utilizing the Seq2Seq model for action prediction

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 human motion prediction method based on attention mechanism
  • A human motion prediction method based on attention mechanism
  • A human motion prediction method based on attention mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0050] A kind of human action prediction method based on attention mechanism, selects Tensorflow to build the attention model that is used for action prediction, comprises the following steps:

[0051] Step 1: According to the instruction, read all the data from the dataset. For example, "eating", iterates through all folders and reads according to the file. In this embodiment, the data set used is the human3.6m data set, so each frame of the read data is divided into 99 float-type values, which are respectively the position information data (three-dimensional) of a root node and the index of 32 parts Map rotation angle.

[0052] Step 2: Standardize the data. First, calculate the standard deviation of all the obtained data according to each column, remove the data with a standard deviation of 0, and obtain all columns with a standard deviation other than 0, subtract each value of each row from the mean value of the column it is in, and then Divide the standard deviation of ...

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 human motion prediction method based on attention mechanism, belonging to the technical field of human-computer interaction and virtual reality. The method of the inventionreads all the data from the human body movement posture data set and performs standardization processing, then reads the training data batch according to the generated random number according to the length of the frame needed to be trained in the instruction, inputs the data into the attention model for data processing, and performs motion reconstruction according to the processed data. The invention adopts two kinds of attention mechanisms, namely global attention mechanism and local attention mechanism. Compared with the prior art, the method of the invention can avoid information redundancyin a certain time and distribute attention under a certain window at the same time.

Description

technical field [0001] The invention relates to a human body action prediction method based on an attention mechanism, which belongs to the technical field of human-computer interaction and virtual reality. Background technique [0002] Humans are predictive, able to make accurate short-term predictions about the world around them based on previous events. In the field of virtual reality, human-computer interaction is an important research direction. How to enable machines to imitate human's ability to make corresponding predictions on human actions is currently a research hotspot in this field. [0003] For computers, predicting human motion is important for timely human-computer interaction, obstacle avoidance, and person tracking. While simple physical phenomena, such as the motion of inanimate objects, can be predicted using known laws of physics, there is no simple equation that governs a person's conscious movement. Many everyday problems, such as predicting what an...

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
IPC IPC(8): G06N3/04G06N3/00
CPCG06N3/008G06N3/045
Inventor 余月陈相儒田聂豪黄焯恒陈宇峰
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
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