Export Reviews, Discussions, Author Feedback and Meta-Reviews

To make the paper stronger, the authors can try to use the STOA Siamese CNN approach of [24] and see how it compares to the proposed methods. Q2: Please summarize your review in 1-2 sentences This paper proposes to use an autoencoder for pose representation learning on faces data. The problems are interesting and the experiments seems to suggest the usefulness of the approach.