Air-writing 인식을 위한 RNN 기법들의 성능 비교

초록

In our work, we use our dataset which is extracted feature from three-dimensional coordinates of the hand and three types of Recurrent Neural Network(RNN). By using our dataset and RNN, we showed that not only which type of RNN is better when using Air-writing recognition but also our method is better than existing Air-writing recognition performance.