Trustworthy Self-Attention: Enabling the Network to Focus Only on the Most Relevant References

arXiv:2403.00211v1 Announce Type: new
Abstract: The prediction of optical flow for occluded points is still a difficult problem that has not yet been solved. Recent methods use self-attention to find relevant non-occluded points as references for estimating the optical flow of occluded points based on the assumption of self-similarity. However, they rely on visual features of a single image and weak constraints, which are not sufficient to constrain the trained network to focus on erroneous and weakly relevant reference points. …

abstract arxiv attention cs.cv enabling flow focus network optical optical flow prediction self-attention trustworthy type