Export Reviews, Discussions, Author Feedback and Meta-Reviews

"NIPS Neural Information Processing Systems 8-11th December 2014, Montreal, Canada",,, "Paper ID:","845" "Title:","Do Convnets Learn Correspondence?" First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. In this paper the authors study the ability of convolutional networks to localize things precisely. The hypothesis is that convolutional networks are able to extract precise location information despite the coarseness of the features/receptive field sizes. The paper presents an interesting new way to visualize the features of the pooling regions in a conv-net, designs an experiment for perfoming intra-class alignment, and studies keypoint classification using conv-net features.