通过四步策略，让模型从不确定中学习，提升小类和模糊区域的分割效果。
FARCLUSS: Fuzzy Adaptive Rebalancing and Contrastive Uncertainty Learning for Semi-Supervised Semantic Segmentation
- 用软伪标签保留多类别概率分布，避免信息丢失
- 按置信度动态加权像素，减少噪声干扰
- 自适应调整损失平衡长尾分布，改善小类表现
半监督语义分割（SSSS）长期面临难以有效利用未标记数据的问题，如伪标签利用率低、类别不平衡加剧以及预测不确定性被忽视。现有方法常通过严格阈值丢弃不确定区域，偏向主导类别。为此，本文提出一个整体框架，将不确定性转化为学习资源，包含四个核心组件：(1) 模糊伪标签，保留top-K预测中的软类别分布以增强监督；(2) 不确定性感知的动态加权，基于熵的可靠性分数调节像素级贡献；(3) 自适应类别重平衡，动态调整损失以缓解长尾分布问题；(4) 轻量级对比正则化，促使特征嵌入紧凑且可区分。在多个基准上的实验表明，该方法显著优于当前最先进方法，在少数类和模糊区域的分割上均有明显提升。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。