Maskomaly:Zero-Shot Mask Anomaly Segmentation
This work addresses the challenge of detecting anomalies in segmentation tasks for applications like autonomous driving, offering a practical solution without needing anomalous data, though it is incremental as it builds on existing segmentation networks.
The authors tackled the problem of anomaly segmentation without requiring anomalous training data by introducing Maskomaly, a framework that adds a simple post-processing step to mask-based semantic segmentation networks, achieving top results on benchmarks like SMIYC, RoadAnomaly, and StreetHazards, with concrete performance improvements such as outperforming all directly comparable approaches on SMIYC.
We present a simple and practical framework for anomaly segmentation called Maskomaly. It builds upon mask-based standard semantic segmentation networks by adding a simple inference-time post-processing step which leverages the raw mask outputs of such networks. Maskomaly does not require additional training and only adds a small computational overhead to inference. Most importantly, it does not require anomalous data at training. We show top results for our method on SMIYC, RoadAnomaly, and StreetHazards. On the most central benchmark, SMIYC, Maskomaly outperforms all directly comparable approaches. Further, we introduce a novel metric that benefits the development of robust anomaly segmentation methods and demonstrate its informativeness on RoadAnomaly.