A neural network based on SPD manifold learning for skeleton-based hand gesture recognition
This work addresses hand gesture recognition for human-computer interaction, presenting an incremental improvement with a novel network architecture.
The paper tackled skeleton-based hand gesture recognition by proposing a neural network that uses SPD manifold learning, achieving state-of-the-art accuracies on two challenging datasets.
This paper proposes a new neural network based on SPD manifold learning for skeleton-based hand gesture recognition. Given the stream of hand's joint positions, our approach combines two aggregation processes on respectively spatial and temporal domains. The pipeline of our network architecture consists in three main stages. The first stage is based on a convolutional layer to increase the discriminative power of learned features. The second stage relies on different architectures for spatial and temporal Gaussian aggregation of joint features. The third stage learns a final SPD matrix from skeletal data. A new type of layer is proposed for the third stage, based on a variant of stochastic gradient descent on Stiefel manifolds. The proposed network is validated on two challenging datasets and shows state-of-the-art accuracies on both datasets.