Skip to content

Real-time Custom Hand Gesture Recognition with PyTorch on EgoGesture, NvGesture, Jester, Kinetics and UCF101

License

Notifications You must be signed in to change notification settings

Solvve/ml_real_time_gesrec

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

59 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Hand gesture recognition

License Python 3.8 scikit-learn 0.23.2 pandas 1.2.3 numpy 1.20.2 torch 0.23.2 Solvve

Description

This is an example of hand recognition gesture task.

The main idea it was add new gesture class without retrain whole model or head layers. For solving this problem we follow the next steps:

  • Remove fc and softmax
  • Generate a vector representation for each gesture in train

Forked from https://github.com/ahmetgunduz/Real-time-GesRec.

To classify vectors we training 3 different classifiers with comparing results and training time:

Classifier Accuracy Time (sec)
KNN 0.89 108
NC + KNN 0.87 1.95
SVC 0.9 1665

Research in notebook -> embedding_representations.ipynb Demo script with camera capture -> simple_online_demo.py

Example demo (54 sec):

About

Real-time Custom Hand Gesture Recognition with PyTorch on EgoGesture, NvGesture, Jester, Kinetics and UCF101

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 96.0%
  • Python 3.7%
  • Shell 0.3%