Learning to Rank Patches for Unbiased Image Redundancy Reduction

arXiv:2404.00680v1 Announce Type: new
Abstract: Images suffer from heavy spatial redundancy because pixels in neighboring regions are spatially correlated. Existing approaches strive to overcome this limitation by reducing less meaningful image regions. However, current leading methods rely on supervisory signals. They may compel models to preserve content that aligns with labeled categories and discard content belonging to unlabeled categories. This categorical inductive bias makes these methods less effective in real-world scenarios. To address this issue, we propose a self-supervised framework …

abstract arxiv cs.cv current however image images pixels redundancy spatial type unbiased