A PyTorch implementation of the DeFoG model for training and sampling discrete graph flows. (Please update to the latest commit. Recent fixes have been applied.) Working with directed graphs? Consider ...

Abstract: The accelerating rate of scientific publications makes it extremely difficult for researchers to find out the relevant papers and related works. Recommender systems that aim at solving the ...

Abstract: Point cloud segmentation is fundamental in many medical applications, such as aneurysm clipping and orthodontic planning. Recent methods mainly focus on designing powerful local feature ...

Previous diffusion models cannot accurately model the graph structures as they learn to denoise at each step without considering the topology of the graphs to be generated. To fix such a myopic ...