Skip to content

Whale-Yu/HandPoseEstimationBasedOnMediapipeAndLstm

Repository files navigation

👇体感互动——Hand Pose Estimation Based on Mediapipe And LSTM

目录:

🐱Introduce

体感互动——Hand Pose Estimation Based on Mediapipe And LSTM

基于mediapipe提取人体姿势、人脸、手部关键点和LSTM算法实现手势识别,并进一步实现体感互动的功能,如进行隔空移动、抓取、放大、缩小等手势操作,可为商业显示提供智能交互应用,如3D模型的展示。

数据:一共采集3类手势,每一个手势30个视频,每一个视频30帧,每一帧提取的landmarks序列长度为1662(33个4维人体姿势landmarks,468个3维人脸landmarks,21个3维左手landmarks,21个3维右手landmarks,33✖4+468✖3+21✖3+21✖3=1662)。

网络:三层lstm(lstm_net.py)

  • demo1:01_predict_hand__pose_.py

    demo1.gif

  • demo2:02_predict.py

    demo2.gif

  • demo3:03_predict之体感互动.py

    demo3.gif

🐖QuickStart

Dependencies

pip install -r requirements.txt

Inference1

python 01_predict_hand_pose.py

Inference2

python 02_predict.py

体感互动

1、 打开https://720.vrqjcs.com/t/9332870054821ffc

2、 python 03_predict之体感互动.py

😔Training

详见train_my_hand_pose/hand pose detection.ipynb

具体开发流程见Hand-Pose-Estimation.pdf 思维导图

🐒Reference

🐕Thanks

  • @Studio:JHC Software Dev Studio

  • @Mentor:HuangRiChen

  • @Author:YuJunYu

About

体感交互——基于Mediapipe和LSTM的手姿估计

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published