this project implements the hand gesture recognition algorithm introduced in paper online detection and classification of dynamic hand gestures with recurrent 3d convolutional neural networks
this project is an implement of dynamic hand gesture recognition algorithm which takes advantages of both C3D and LSTM. because part of the model is adopted from C3D, the model is finetuned from trained models from C3D which is provided from another repo of my git.
download the dataset here and unzip it under the current directory. then use the following command to generate dataset.
python create_dataset.pyplace the trained C3D model directory under current directory and train with command
python train_r3dcnn.pytest on video captured from webcam with command
python ActionRecognition.py