NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:827
Title:Neural Diffusion Distance for Image Segmentation

The authors propose a combination of diffusion distances with deep networks trained for image segmentation. Two of the reviewers felt that there were interesting methodological contributions meriting acceptance, while all three reviewers indicated that baselines in the experiments could be improved.