Thank you for releasing the depth2image model. I wonder why you prefer depth image over edge image. I think edge image can also provide shape information. Is it because depth image is less sparse? It would be great if you could share your thoughts on it.
