Fuzzy Uncertainty-Based Out-of-Distribution Detection Algorithm for Semantic Segmentation

Lin, Qiao; Chen, Xin; Chen, Chao; Pekaslan, Direnc; Garibaldi, Jonathan M.

Abstract

Deep learning models have achieved high performance in numerous semantic segmentation tasks. However, when the input data at test time do not resemble the training data, deep learning models can not handle them properly and will probably produce poor results. Therefore, it is important to design algorithms for deep learning models to reliably detect out-of-distribution (OOD) data. In this paper, we propose a novel fuzzy-uncertainty-based method to detect OOD samples for semantic segmentation. Firstly, to capture both data and model uncertainties, test-time augmentation and Monte Carlo dropout are applied to a ready-trained image segmentation model for generating multiple predicted instances of a given test image. Then interval fuzzy sets are generated from these multiple predictions to describe the captured uncertainty via distance transform operators. Finally, an image-level uncertainty score, which is calculated from the generated interval fuzzy sets, is used to indicate if it is an OOD sample. Experiments on testing three OOD test sets on a skin lesion segmentation model show that our proposed method achieved significantly higher classification accuracy in detecting OOD samples than three other state-of-the-art uncertainty-based algorithms.

Citation

Lin, Q., Chen, X., Chen, C., Pekaslan, D., & Garibaldi, J. M. (2023, August). Fuzzy Uncertainty-Based Out-of-Distribution Detection Algorithm for Semantic Segmentation. Presented at 2023 IEEE International Conference on Fuzzy Systems (FUZZ), Songdo Incheon, Korea

• Presentation Conference Type: Edited Proceedings
• Conference Name: 2023 IEEE International Conference on Fuzzy Systems (FUZZ)
• Start Date: Aug 13, 2023
• End Date: Aug 17, 2023
• Acceptance Date: Aug 13, 2023
• Online Publication Date: Nov 9, 2023
• Publication Date: Aug 13, 2023
• Deposit Date: Nov 5, 2024
• Publicly Available Date: Dec 20, 2024
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Pages: 115-120
• Book Title: 2023 IEEE International Conference on Fuzzy Systems (FUZZ) Proceedings
• ISBN: 979-8-3503-3229-2
• DOI: https://doi.org/10.1109/fuzz52849.2023.10309696
• Public URL: https://nottingham-repository.worktribe.com/output/27583553
• Publisher URL: https://ieeexplore.ieee.org/document/10309696