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

Massage manipulation recognition method based on deep learning

A technology of deep learning and recognition methods, applied in neural learning methods, character and pattern recognition, force/torque/power measuring instruments, etc., can solve problems such as inability to collect massage force information, inability to extract timing information, and a large amount of cost. Achieve the effect of short network training time, improved recognition accuracy, and high recognition accuracy

Pending Publication Date: 2022-03-11
SHANGHAI UNIV
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the visual sensor can only collect the movement information of the massage technique, but not the force information of the massage technique; the two-dimensional convolutional neural network can only be trained to extract the spatial domain information of the massage technique, but cannot extract the time series information closely related to the massage technique ; In addition, traditional sensor data acquisition also requires a lot of cost

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
  • Massage manipulation recognition method based on deep learning
  • Massage manipulation recognition method based on deep learning
  • Massage manipulation recognition method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] see figure 1 , a massage technique recognition method based on deep learning, the operation steps are as follows:

[0045] Step 1: Collect the data corresponding to the massage action through the flexible distributed tactile sensor;

[0046] Step 2: Visualize the sensor data as a massage dot matrix heat map through the host computer;

[0047] Step 3: Construct a model that generates target data from hidden variables through variational autoencoder VAE to expand the original collected data;

[0048] Step 4: Extract the key frame of the massage lattice heat map through the frame difference method;

[0049] Step 5: Use a two-dimensional convolutional neural network to extract the spatial features of the input massage force bitmaps of each frame;

[0050] Step 6: Introduce the frame attention mechanism after the convolutional neural network, and assign weight values ​​to the video frame dimension of the data;

[0051] Step 7: use the cyclic neural network to extract the...

Embodiment 2

[0055] This embodiment is basically the same as Embodiment 1, and the special features are as follows:

[0056] In the present embodiment, in the step 2, the massage data corresponding to the collected massage action is visualized, and each sensing unit of the tactile sensor is visualized in the host computer according to its actual distribution position on the glove through Matlab. The heat map shows the magnitude of the force on the sensing unit, and the force on the sensing unit is from small to large, corresponding to the color of the sensing unit from cold to warm.

[0057] In the third step, the variational autoencoder VAE indirectly obtains the distribution of real sample data by introducing hidden variables, interpolates the hidden variables and decodes them into generated samples, and expands massage points without additional data collection Array heat map to achieve data enhancement.

[0058] In the fourth step, the frame difference method averages the pixels of eve...

Embodiment 3

[0065] In this example, if figure 1 As shown in the flow chart of this embodiment, the system is divided into three major modules: a data acquisition module, a data processing module, and a deep learning module, specifically including the following steps:

[0066] Step 1: In the data acquisition module, the magnitude and distribution of the force exerted by the hands of the doctor during massage are collected through the tactile sensor and the data acquisition system; and the massage data acquisition is realized.

[0067] Step 2: In the data processing module, the upper computer receives data in real time through Matlab and converts the data in ASCII format into plastic format, and then visualizes the data as a massage lattice heat map and saves it. A massage action corresponds to multiple frames arranged in time order Massage lattice heat map; realize data visualization.

[0068] Step 3: In the data processing module, the data uses a variational auto-encoder (VAE) to constru...

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 massage manipulation recognition method based on deep learning, and the method comprises the steps: collecting the force distribution and force size information of a hand during massage through a flexible touch sensor, extracting the manipulation features through a neural network, and achieving the recognition of a massage manipulation. A variational auto-encoder is adopted to realize data enhancement; extracting key frames of input data by using a frame difference method, and removing input redundant frames; extracting and training spatial domain and time domain features of the massage dot matrix thermodynamic diagram group through a two-dimensional convolutional neural network and a recurrent neural network; a frame attention mechanism is introduced after the convolutional neural network, and the massage manipulation recognition precision of the network is improved. According to the invention, the original sensor data is expanded without increasing the data acquisition cost; key frames of graph group data are extracted, the network overfitting phenomenon is reduced, and the network generalization ability is improved; the neural network extracts time domain information between the video frames to obtain time domain features of the massage manipulation; and by introducing a frame attention mechanism, the recognition precision is effectively improved.

Description

technical field [0001] The invention belongs to the field of artificial intelligence, and in particular relates to a massage technique recognition method based on deep learning. Background technique [0002] Massage is a method of acting on specific parts of the human body surface with manipulations to adjust the physiological and pathological conditions of the body and achieve the purpose of physical therapy. Traditional manual massage requires a lot of physical strength of the masseur, and also requires a lot of training costs. [0003] With the development of robotics, robots can replace human beings in massage services. In order to realize robot massage, it is first necessary to systematically understand the massage techniques of professional masseurs, explore the characteristics of massage techniques, and provide a reference for robots to reproduce massage techniques of masseurs. [0004] At present, domestic and foreign manipulation recognition mostly uses visual sen...

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): G06V40/20G06V10/82G06N3/04G06N3/08G01L5/00
CPCG06N3/084G01L5/00G06N3/045G06N3/044
Inventor 雷静桃朱盛鼎陈冬冬
Owner SHANGHAI 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