Conformal Semantic Image Segmentation: Post-hoc Quantification of Predictive Uncertainty

arXiv:2405.05145v1 Announce Type: cross
Abstract: We propose a post-hoc, computationally lightweight method to quantify predictive uncertainty in semantic image segmentation. Our approach uses conformal prediction to generate statistically valid prediction sets that are guaranteed to include the ground-truth segmentation mask at a predefined confidence level. We introduce a novel visualization technique of conformalized predictions based on heatmaps, and provide metrics to assess their empirical validity. We demonstrate the effectiveness of our approach on well-known benchmark datasets and image segmentation prediction …

abstract arxiv confidence cs.cv cs.lg generate ground-truth image novel prediction predictive quantification segmentation semantic truth type uncertainty visualization