FSNet: A Failure Detection Framework for Semantic Segmentation
This addresses safety-critical failure detection in autonomous driving, but it is incremental as it builds on existing failure detection methods with specific gains.
The paper tackles the problem of detecting pixel-level misclassifications in semantic segmentation models for autonomous vehicles, achieving performance improvements of 12.30%, 9.46%, and 9.65% in AUPR-Error on Cityscapes, BDD100K, and Mapillary datasets.
Semantic segmentation is an important task that helps autonomous vehicles understand their surroundings and navigate safely. During deployment, even the most mature segmentation models are vulnerable to various external factors that can degrade the segmentation performance with potentially catastrophic consequences for the vehicle and its surroundings. To address this issue, we propose a failure detection framework to identify pixel-level misclassification. We do so by exploiting internal features of the segmentation model and training it simultaneously with a failure detection network. During deployment, the failure detector can flag areas in the image where the segmentation model have failed to segment correctly. We evaluate the proposed approach against state-of-the-art methods and achieve 12.30%, 9.46%, and 9.65% performance improvement in the AUPR-Error metric for Cityscapes, BDD100K, and Mapillary semantic segmentation datasets.