FARCLUSS: Fuzzy Adaptive Rebalancing and Contrastive Uncertainty Learning for Semi-Supervised Semantic Segmentation
By: Ebenezer Tarubinga, Jenifer Kalafatovich, Seong-Whan Lee
Potential Business Impact:
Teaches computers to see better, even in tricky spots.
Semi-supervised semantic segmentation (SSSS) faces persistent challenges in effectively leveraging unlabeled data, such as ineffective utilization of pseudo-labels, exacerbation of class imbalance biases, and neglect of prediction uncertainty. Current approaches often discard uncertain regions through strict thresholding favouring dominant classes. To address these limitations, we introduce a holistic framework that transforms uncertainty into a learning asset through four principal components: (1) fuzzy pseudo-labeling, which preserves soft class distributions from top-K predictions to enrich supervision; (2) uncertainty-aware dynamic weighting, that modulate pixel-wise contributions via entropy-based reliability scores; (3) adaptive class rebalancing, which dynamically adjust losses to counteract long-tailed class distributions; and (4) lightweight contrastive regularization, that encourage compact and discriminative feature embeddings. Extensive experiments on benchmarks demonstrate that our method outperforms current state-of-the-art approaches, achieving significant improvements in the segmentation of under-represented classes and ambiguous regions.
Similar Papers
Rethinking Semi-supervised Segmentation Beyond Accuracy: Reliability and Robustness
CV and Pattern Recognition
Makes self-driving cars safer by checking their vision.
When Confidence Fails: Revisiting Pseudo-Label Selection in Semi-supervised Semantic Segmentation
CV and Pattern Recognition
Teaches computers to better understand pictures.
Uncertainty-Supervised Interpretable and Robust Evidential Segmentation
CV and Pattern Recognition
Makes medical scans more trustworthy and reliable.