SemPLeS: Semantic Prompt Learning for Weakly-Supervised Semantic Segmentation

arXiv:2401.11791v2 Announce Type: replace-cross
Abstract: Weakly-Supervised Semantic Segmentation (WSSS) aims to train segmentation models using image data with only image-level supervision. Since precise pixel-level annotations are not accessible, existing methods typically focus on producing pseudo masks for training segmentation models by refining CAM-like heatmaps. However, the produced heatmaps may capture only the discriminative image regions of object categories or the associated co-occurring backgrounds. To address the issues, we propose a Semantic Prompt Learning for WSSS (SemPLeS) framework, which learns to …

abstract annotations arxiv cs.cl cs.cv cs.lg data focus however image image data masks pixel prompt prompt learning segmentation semantic supervision train training type weakly-supervised