Guangming Shi

TL;DR: A novel spatial-temporal Gaussian scale mixture (STGSM) model for foreground estimation is proposed, and the optical flow has been used to model the correspondences between foreground pixels in adjacent frames to better characterize the temporal correlations.

TL;DR: This paper addresses the challenging unsupervised scene flow estimation problem by jointly learning four low-level vision sub-tasks: optical flow F, stereo-depth D, camera pose P and motion segmentation S by designing a novel Rigidity From Motion (RfM) layer with three principal components.

TL;DR: In this article, the authors proposed a federated edge intelligence (FEI) framework that allows edge servers to evaluate the required number of data samples according to the energy cost of the IoT network as well as their local data processing capacity and only request the amount of data that is sufficient for training a satisfactory model.

TL;DR: This paper proposes a model inspired by the observations in three-dimensional environment to better present the influence of depth, and demonstrates that the proposed method can outperform the state-of-the-art fixation prediction algorithms on several public data sets for stereoscopic saliency estimation.

TL;DR: Numerical experiments show that the proposed method not only has well anti-attack ability but also is robust to packet loss, which can still decrypt plain-image even when the packet loss ratio is up to 50%.