홈 › Papers

The Impact of CutMix on Reliability and Robustness in Semantic Segmentation

Ensuring not only high accuracy but also reliable and robust predictions is critical for the deployment of semantic segmentation models in safety-critical applications such as autonomous driving. Despite the widespread use of CutMix - a simple yet powerful data augmentation strategy - its effect on the reliability and robustness in dense predictions tasks remains unexplored. Motivated by recent findings that semi-supervised segmentation methods, where CutMix is a core component, can severely degrade reliability, this study isolates and systematically analyzes the influence of CutMix on segmentation accuracy, calibration, and uncertainty quality. We evaluate two representative architectures, the CNN-based DeepLabV3+ and the transformer-based SegFormer, across both in-domain and out-of-domain scenarios. Our results show that CutMix has only a minor impact on segmentation accuracy but consistently improves the reliability, particularly under distribution shifts. These improvements indicate that CutMix primarily enhances the trustworthiness of the model's calibration and uncertainty rather than the raw segmentation prediction itself. This distinction is crucial for safety-critical deployment, where reliable confidence estimates are as important as raw performance.

Code (0)

등록된 구현이 없습니다.

Tasks

Similar Papers 제목 키워드 기반