The network is a differentiable deep architecture consisting of feature extraction and diffusion distance modules for computing diffusion distance on image by end-to-end training. We design low resolution kernel matching loss and high resolution segment matching loss to enforce the network's output to beconsistent withhuman-labeled image segments.

Extended " T able 2" in submitted paper. Extended " T able 3" in submitted paper. We thank reviewers for their comments, and will carefully revise paper considering these comments. Q1 (R1): References and comparison with a baseline that learns embeddings only through a standard convnet. In Tab.2 of this rebuttal, the state-of-the-art method of AISI [7] also depends on We will give more details of these compared methods in paper for clarity.

The network is a differentiable deep architecture consisting of feature extraction and diffusion distance modules for computing diffusion distance on image by end-to-end training. We design low resolution kernel matching loss and high resolution segment matching loss to enforce the network's output to be consistent with human-labeled image segments.

Extended " T able 2" in submitted paper. Extended " T able 3" in submitted paper. We thank reviewers for their comments, and will carefully revise paper considering these comments. Q1 (R1): References and comparison with a baseline that learns embeddings only through a standard convnet. In Tab.2 of this rebuttal, the state-of-the-art method of AISI [7] also depends on We will give more details of these compared methods in paper for clarity.

The paper is clearly written, everything proposed in the paper makes sense and seems like a natural thing to do (I had been working on the same problem, so I am entirely in favor of the pursued direction). This is not only important for giving credit to earlier works. A more crucial question in connection with these works, is whether the structured layer adds something on top of the ability of a cnn to compute embeddings for image segmentation. In experiments that I have been working on it has been really hard to beat a well-tuned, plain convnet trained with a siamese loss, and introducing a spectral normalization layer only added complications. It would be really useful if the authors could do this comparison on top of a strong baseline (e.g. the methods mentioned above) and indicate whether the resulting embeddings ( eigenvectors) are any better than those delivered from the original baselines.

Diffusion distance is a spectral method for measuring distance among nodes on graph considering global data structure. In this work, we propose a spec-diff-net for computing diffusion distance on graph based on approximate spectral decomposition. The network is a differentiable deep architecture consisting of feature extraction and diffusion distance modules for computing diffusion distance on image by end-to-end training. We design low resolution kernel matching loss and high resolution segment matching loss to enforce the network's output to be consistent with human-labeled image segments. To compute high-resolution diffusion distance or segmentation mask, we design an up-sampling strategy by feature-attentional interpolation which can be learned when training spec-diff-net.

Diffusion distance is a spectral method for measuring distance among nodes on graph considering global data structure. In this work, we propose a spec-diff-net for computing diffusion distance on graph based on approximate spectral decomposition. The network is a differentiable deep architecture consisting of feature extraction and diffusion distance modules for computing diffusion distance on image by end-to-end training. We design low resolution kernel matching loss and high resolution segment matching loss to enforce the network's output to be consistent with human-labeled image segments. To compute high-resolution diffusion distance or segmentation mask, we design an up-sampling strategy by feature-attentional interpolation which can be learned when training spec-diff-net.