Dynamic Hand Gesture Recognition for Wearable Devices with Low Complexity Recurrent Neural Networks
This work addresses gesture recognition for wearable devices, offering incremental improvements in efficiency through quantization.
The paper tackles dynamic hand gesture recognition for wearable devices by developing two low-complexity recurrent neural network (RNN) techniques, one using video with CNN-RNN and another using accelerometer data with RNN, achieving optimization through fixed-point quantization of weights to two bits to reduce memory and power consumption.
Gesture recognition is a very essential technology for many wearable devices. While previous algorithms are mostly based on statistical methods including the hidden Markov model, we develop two dynamic hand gesture recognition techniques using low complexity recurrent neural network (RNN) algorithms. One is based on video signal and employs a combined structure of a convolutional neural network (CNN) and an RNN. The other uses accelerometer data and only requires an RNN. Fixed-point optimization that quantizes most of the weights into two bits is conducted to optimize the amount of memory size for weight storage and reduce the power consumption in hardware and software based implementations.