Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Vorndran, Michael R. H., Roeck, Bernhard F.
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2401.14387
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910000213917696
author Vorndran, Michael R. H.
Roeck, Bernhard F.
author_facet Vorndran, Michael R. H.
Roeck, Bernhard F.
contents A primary challenge in semi-supervised learning (SSL) for segmentation is the confirmation bias from noisy pseudo-labels, which destabilizes training and degrades performance. We propose Inconsistency Masks (IM), a framework that reframes model disagreement not as noise to be averaged away, but as a valuable signal for identifying uncertainty. IM leverages an ensemble of teacher models to generate a mask that explicitly delineates regions where predictions diverge. By filtering these inconsistent areas from input-pseudo-label pairs, our method effectively mitigates the cycle of error propagation common in both continuous and iterative self-training paradigms. Extensive experiments on the Cityscapes benchmark demonstrate IM's effectiveness as a general enhancement framework: when paired with leading approaches like iMAS, U$^2$PL, and UniMatch, our method consistently boosts accuracy, achieving superior benchmarks across ResNet-50 and DINOv2 backbones, and even improving distilled architectures like SegKC. Furthermore, the method's robustness is confirmed in resource-constrained scenarios where pre-trained weights are unavailable. On three additional diverse datasets from medical and underwater domains trained entirely from scratch, IM significantly outperforms standard SSL baselines. Notably, the IM framework is dataset-agnostic, seamlessly handling binary, multi-class, and complex multi-label tasks by operating on discretized predictions. By prioritizing training stability, IM offers a generalizable and robust solution for semi-supervised segmentation, particularly in specialized areas lacking large-scale pre-training data. The full code is available at: https://github.com/MichaelVorndran/InconsistencyMasks
format Preprint
id arxiv_https___arxiv_org_abs_2401_14387
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Inconsistency Masks: Harnessing Model Disagreement for Stable Semi-Supervised Segmentation
Vorndran, Michael R. H.
Roeck, Bernhard F.
Computer Vision and Pattern Recognition
A primary challenge in semi-supervised learning (SSL) for segmentation is the confirmation bias from noisy pseudo-labels, which destabilizes training and degrades performance. We propose Inconsistency Masks (IM), a framework that reframes model disagreement not as noise to be averaged away, but as a valuable signal for identifying uncertainty. IM leverages an ensemble of teacher models to generate a mask that explicitly delineates regions where predictions diverge. By filtering these inconsistent areas from input-pseudo-label pairs, our method effectively mitigates the cycle of error propagation common in both continuous and iterative self-training paradigms. Extensive experiments on the Cityscapes benchmark demonstrate IM's effectiveness as a general enhancement framework: when paired with leading approaches like iMAS, U$^2$PL, and UniMatch, our method consistently boosts accuracy, achieving superior benchmarks across ResNet-50 and DINOv2 backbones, and even improving distilled architectures like SegKC. Furthermore, the method's robustness is confirmed in resource-constrained scenarios where pre-trained weights are unavailable. On three additional diverse datasets from medical and underwater domains trained entirely from scratch, IM significantly outperforms standard SSL baselines. Notably, the IM framework is dataset-agnostic, seamlessly handling binary, multi-class, and complex multi-label tasks by operating on discretized predictions. By prioritizing training stability, IM offers a generalizable and robust solution for semi-supervised segmentation, particularly in specialized areas lacking large-scale pre-training data. The full code is available at: https://github.com/MichaelVorndran/InconsistencyMasks
title Inconsistency Masks: Harnessing Model Disagreement for Stable Semi-Supervised Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2401.14387