Houqiang Li

TL;DR: In this article, a novel passive non-line-of-sight (NLOS) imaging framework based on manifold embedding and optimal transport is proposed to reconstruct high-quality complicated hidden scenes.

TL;DR: Experiments on SVHN, CIFAR and ILSVRC-2012 datasets prove that the proposed method can bring faster convergence and higher accuracy for low-precision neural networks.

TL;DR: It is shown that this list secrecy problem is equivalent to the one with secrecy measured by a new quantity lossy equivocation, which is proved to be the minimum optimistic one-achievable source coding rate, and the achievable region for Gaussian communication case.

TL;DR: Wang et al. as discussed by the authors proposed a novel advanced motion model to handle the irregular motion for the cubic map projection of 360-degree video, where they first try to project the pixels in both the current picture and reference picture from unfolding cube back to the sphere, then through utilizing the characteristic that most of the motions in the sphere are uniform, they can derive the relationship between the motion vectors of various pixels in the unfold cube.

TL;DR: In this article, a neural-network-based method for cross-channel prediction in intra frame coding is proposed, which utilizes twofold cues, i.e., the neighboring reconstructed samples with all channels, and the co-located reconstructed sampled with partial channels.