NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:403
Title:Volumetric Correspondence Networks for Optical Flow

• This paper presents a method for dense optical flow estimation Proposing a 4D method capable to reduce the significant amount of memory. The proposal is very efficient in term of speed and it will be very useful in computer vision. • The three reviewers were not concordant in term of rate ( 1 reject and two accept rates) • After the rebuttal the two positive reviewers were satisfied by the rebuttal keeping positive the rate. The first reviewer still was not convinced about the novelty and the clearness of the method. After a discussion the area chair suggest an acceptance.