Ding Yuan

TL;DR: Inspired by the Gaussian-like shape of the small target, the local intensity and gradient map is calculated from the original infrared image in order to enhance the targets and suppress clutters.

TL;DR: A simple but effective image prior, change of detail (CoD) prior, to remove haze from a single input image, which can be implemented very quickly and stable to image local regions containing objects in different depths.

TL;DR: Performance evaluation on the 27 Middlebury data sets shows that both the algorithms outperform the other two tree-based non-local methods, namely minimum spanning tree (MST) and segment-tree (ST).

TL;DR: This paper proposes a learning framework using static, dynamic and sequential mixed features to solve three fundamental problems: spatial domain variation, temporal domain polytrope, and intra- and inter-class diversities to extract human action representations which are robust against spatial and temporal interferences.

TL;DR: The proposed matching cost outperforms traditional matching costs, the discriminative dictionary learning model is more suitable than previous dictionary learning models for stereo matching, and the proposed stereo method ranks the third place on the Middlebury benchmark v3 in quarter resolution up to the submitting, and achieves the best accuracy on 30 classic stereo images.