Exploiting LSTM-RNNs and 3D Skeleton Features for Hand Gesture Recognition
Heyuan Guo, Yang Yang, Hua Cai · 2019
With the development of deep learning, a hand gesture recognition solution using LSTM-RNNs and 3D Skeleton Features is presented. The 3D skeleton data is acquired by the Kinect sensor. According to the relevant skeleton information, feature vectors are constructed; These hand gestures can be represented as sets of feature vectors that change over time. Recurrent Neural Networks (RNNs) are suited to analyse this type of sets thanks to their ability to model the long term contextual information of temporal sequences. The LSTM is an architecture where RNNs use special units instead of common activation function. Finally, the proposed method was evaluated on the NTU RGB+D dataset. The experimental results show that the proposed method has an accuracy of 92.196% on the selfdefined dataset and it has good robustness.