Siming Yan

TL;DR: Recently, this article showed that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that of models derived using today's best supervised methods and that the mapping of neural network hidden layers is neuroanatomically consistent across the ventral stream.

TL;DR: It is found that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that of models derived using today’s best supervised methods.

TL;DR: Implicit Autoencoder (IAE) is introduced, a simple yet effective method that addresses the challenge of autoencoding on point clouds by replacing the point cloud decoder with an implicit decoder that outputs a continuous representation that is shared among different point cloud sampling of the same model.

TL;DR: A novel RGB-D based relative pose estimation approach that is suitable for small-overlapping or non- overlapping scans and can output multiple relative poses and considerably boosts the performance of multi-scan reconstruction in few-view reconstruction settings.

TL;DR: HPNet as discussed by the authors leverages hybrid representations that combine one learned semantic descriptor, two spectral descriptors derived from predicted geometric parameters, as well as an adjacency matrix that encodes sharp edges.