Rethinking Saliency-Guided Weakly-Supervised Semantic Segmentation
This work addresses the need for standardized evaluation in weakly-supervised semantic segmentation, which is incremental as it builds on existing methods by providing a framework rather than a new model.
This paper tackles the problem of saliency-guided weakly-supervised semantic segmentation by identifying that the arbitrary choice of saliency maps and thresholds critically impacts performance, and it introduces WSSS-BED, a standardized framework with various saliency and activation maps for seven methods to facilitate more rigorous research.
This paper presents a fresh perspective on the role of saliency maps in weakly-supervised semantic segmentation (WSSS) and offers new insights and research directions based on our empirical findings. We conduct comprehensive experiments and observe that the quality of the saliency map is a critical factor in saliency-guided WSSS approaches. Nonetheless, we find that the saliency maps used in previous works are often arbitrarily chosen, despite their significant impact on WSSS. Additionally, we observe that the choice of the threshold, which has received less attention before, is non-trivial in WSSS. To facilitate more meaningful and rigorous research for saliency-guided WSSS, we introduce \texttt{WSSS-BED}, a standardized framework for conducting research under unified conditions. \texttt{WSSS-BED} provides various saliency maps and activation maps for seven WSSS methods, as well as saliency maps from unsupervised salient object detection models.