SinSim：Sinkhorn 正则化的 SimCLR
SinSim: Sinkhorn-Regularized SimCLR
AI总结：
本文提出 SinSim，将最优传输中的 Sinkhorn 正则化引入 SimCLR，以构建更具几何结构的特征空间，并在多个数据集上验证其优于或比肩主流自监督方法。
摘要：
自监督学习通过消除对标注数据的需求，彻底改变了表征学习。诸如 SimCLR 之类的对比学习方法会最大化同一图像不同增强视图之间的一致性，但缺乏用于强制形成全局结构化潜在空间的显式正则化。这一局限通常会导致次优的泛化性能。我们提出 SinSim，这是 SimCLR 的一种新颖扩展，它融合了来自最优传输理论的 Sinkhorn 正则化，以增强表征结构。Sinkhorn 损失是一种熵正则化的 Wasserstein 距离，可促进分布良好且具备几何感知能力的特征空间，同时保留判别能力。在多个数据集上的实证评估表明，SinSim 的表现优于 SimCLR，并且相对于 VICReg 和 Barlow Twins 等著名自监督方法取得了具有竞争力的性能。UMAP 可视化进一步揭示了更好的类别可分性和结构化的特征分布。这些结果表明，将最优传输正则化集成到对比学习中，为学习鲁棒且结构良好的表征提供了一种有原则且有效的机制。我们的发现为在自监督学习框架中应用基于传输的约束开辟了新方向。
英文摘要：
Self-supervised learning has revolutionized representation learning by eliminating the need for labeled data. Contrastive learning methods, such as SimCLR, maximize the agreement between augmented views of an image but lack explicit regularization to enforce a globally structured latent space. This limitation often leads to suboptimal generalization. We propose SinSim, a novel extension of SimCLR that integrates Sinkhorn regularization from optimal transport theory to enhance representation structure. The Sinkhorn loss, an entropy-regularized Wasserstein distance, encourages a well-dispersed and geometry-aware feature space, preserving discriminative power. Empirical evaluations on various datasets demonstrate that SinSim outperforms SimCLR and achieves competitive performance against prominent self-supervised methods such as VICReg and Barlow Twins. UMAP visualizations further reveal improved class separability and structured feature distributions. These results indicate that integrating optimal transport regularization into contrastive learning provides a principled and effective mechanism for learning robust, well-structured representations. Our findings open new directions for applying transport-based constraints in self-supervised learning frameworks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
