For recognising hand gestures using RNN and LSTM... Implementation in TensorFlow
In this git repo, there are 4 main files that represent the current implementation. Rest are for testing purposes. These 4 files can be copied for other implementations.
-
rnn_lstm_optimized.py --- This python script is the optimized implementation. It trains the model on the data and then tests the accuracy and records it.
-
rnn_lstm_train.py --- This python script file is for training the model on the data and then string the model and the graph in the folder trained_model.
-
rnn_lstm_test.py --- This python script file is for testing the accuracy of the saved trained model.
-
rnn_lstm_testing.py --- It calculates and depicts the accuracy of prediction of the model across the 4 classes of gestures. It also represents the standard deviation for each class during prediction.
Link to the Research Paper --- https://link.springer.com/content/pdf/10.1007%2F978-3-319-72038-8_3.pdf
Please cite the following paper:
Dynamic Hand Gesture Recognition for Mobile Systems Using Deep LSTM (Link for Paper -- https://link.springer.com/content/pdf/10.1007%2F978-3-319-72038-8_3.pdf)
Citation ---
Sarkar, A., Gepperth, A., Handmann, U., & Kopinski, T. (2017, December). Dynamic Hand Gesture Recognition for Mobile Systems Using Deep LSTM. In International Conference on Intelligent Human Computer Interaction (pp. 19-31). Springer, Cham.