NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:2401
Title:Learning Deep Bilinear Transformation for Fine-grained Image Representation

The paper received mixed reviews because reviewers were confused about the method and novelty. After the rebuttal, two reviewers better understood and appreciated the results, as well as difference to the state-of-the-art. Two reviewers argue the novelty is limited. While the novelty is minor, there is novelty, which is articulated by the rebuttal and appreciated by the most favorable reviewer. The paper is accepted on the condition the information from the rebuttal is integrated into the camera ready.