Motion Feature Augmented Recurrent Neural Network for Skeleton-based Dynamic Hand Gesture Recognition

TL;DR

This paper introduces a motion feature augmented RNN that combines finger and global hand motion features with skeleton data to enhance the accuracy of skeleton-based dynamic hand gesture recognition.

Contribution

The paper proposes a novel method that integrates finger and global motion features into a bidirectional RNN for improved gesture recognition accuracy.

Findings

The method outperforms existing state-of-the-art approaches.

Motion features significantly improve recognition performance.

The approach is effective across various gesture datasets.

Abstract

Dynamic hand gesture recognition has attracted increasing interests because of its importance for human computer interaction. In this paper, we propose a new motion feature augmented recurrent neural network for skeleton-based dynamic hand gesture recognition. Finger motion features are extracted to describe finger movements and global motion features are utilized to represent the global movement of hand skeleton. These motion features are then fed into a bidirectional recurrent neural network (RNN) along with the skeleton sequence, which can augment the motion features for RNN and improve the classification performance. Experiments demonstrate that our proposed method is effective and outperforms start-of-the-art methods.