Chenyang Li

TL;DR: The LPSNet is proposed, an end-to-end deep neural network based hand gesture recognition framework with novel log path signature features and a new method based on PS and LPS to effectively combine RGB and depth videos.

TL;DR: Zhang et al. as mentioned in this paper leverage a robust feature descriptor, path signature (PS), and propose three PS features to explicitly represent the spatial and temporal motion characteristics, i.e., spatial PS (S PS), temporal PS (T PS) and temporal spatial PS(T S PS).

TL;DR: A heatmap-based FCN (Fully Convolution Network) named YOLSE (You Only Look what You Should See) for fingertip detection in the egocentric vision from single RGB image is proposed and the fingermap is the proposed new probabilistic representation for the multiple fingertip Detection.

TL;DR: In this article, a first-perspective dynamic gesture recognition method based on a deep convolutional neural network framework was proposed. And the method comprises the steps that digital gesture pictures under different complicated backgrounds are collected, identical gestures have the same tag, and an exterior rectangle of the gestures in the digital gestures is marked out; a deep CNN extracted a plurality of candidate boxes from the digital gesture images, the candidate boxes and the exterior rectangle are subjected to feature comparison, the proposal boxes with intact gestures are saved, feature information in the saved candidate boxes are extracted, an obtained

TL;DR: A robust feature descriptor, path signature (PS), is leveraged, and three PS features are proposed to explicitly represent the spatial and temporal motion characteristics, i.e., spatial PS (S_PS), temporal PS (T_PS) and temporal spatial PS(T_S-PS), which achieves the state-of-the-art performance on skeleton-based gesture recognition with high computational efficiency.