Efficacy of Pixel-Level OOD Detection for Semantic Segmentation
This work addresses a safety-critical need for autonomous driving by enabling pixel-level anomaly detection, but it is incremental as it adapts existing methods to a new task with limited success.
The paper tackled the problem of localizing unusual objects in safety-critical applications like autonomous driving by adapting image-level out-of-distribution detection methods to pixel-level detection for semantic segmentation, finding that performance rankings did not transfer and all methods performed significantly worse than their image-level counterparts.
The detection of out of distribution samples for image classification has been widely researched. Safety critical applications, such as autonomous driving, would benefit from the ability to localise the unusual objects causing the image to be out of distribution. This paper adapts state-of-the-art methods for detecting out of distribution images for image classification to the new task of detecting out of distribution pixels, which can localise the unusual objects. It further experimentally compares the adapted methods on two new datasets derived from existing semantic segmentation datasets using PSPNet and DeeplabV3+ architectures, as well as proposing a new metric for the task. The evaluation shows that the performance ranking of the compared methods does not transfer to the new task and every method performs significantly worse than their image-level counterparts.