学习流向自监督表示
Learning a Flow to Self-Supervised Representations
浏览论文内容
中文总结 AI 辅助
提出非对抗性FBDM框架，通过球形条件速度回归学习自监督表示，性能接近DM，训练速度提升1.48-1.83倍，并提供理论保证。
中文摘要 AI 辅助
显式几何参考为构建自监督表示提供了一种直接方式。然而，现有的对抗性分布匹配公式需要昂贵的编码器-评论家优化。我们引入了基于流的分布匹配（FBDM），这是一种非对抗性框架，通过球形条件速度回归来学习这种参考导向的几何结构。受ETF启发的参考允许其组件数量K'超过辅助流维度d*，同时保持结构化的几何分离。我们将每个图像的两种增强视图分配给同一目标，同时限制每个参考中心可以接收的图像数量。一个显式的对齐损失进一步拉近两个视图的表示。从CIFAR到ImageNet的基准实验表明，FBDM的性能几乎与DM相当，并且与现有的SSL方法保持竞争力。匹配的训练成本比较显示，与DM相比，FBDM实现了1.48倍至1.83倍的加速，而GPU内存使用量增加可忽略不计。我们还对学习到的表示的有用性提供了理论解释：在所述条件下，我们根据FBDM预训练损失限定了下游误分类率。
英文摘要
Explicit geometric references offer a direct way to structure self-supervised representations. Existing adversarial distribution-matching formulations, however, require costly encoder-critic optimization. We introduce Flow-Based Distribution Matching (FBDM), a non-adversarial framework that learns this reference-directed geometry through spherical conditional velocity regression. An ETF-inspired reference allows its number of components K' to exceed the auxiliary flow dimension d* while retaining structured geometric separation. We assign both augmented views of each image to the same target, while limiting how many images each reference center can receive. An explicit alignment loss further pulls the two views' representations closer together. Experiments across benchmarks ranging from CIFAR to ImageNet show that FBDM achieves performance nearly on par with DM and remains competitive with existing SSL methods. Matched training-cost comparisons show a 1.48- to 1.83-fold speedup over DM with a negligible increase in GPU memory usage. We also provide a theoretical explanation for the usefulness of the learned representations: under stated conditions, we bound the downstream misclassification rate in terms of the FBDM pretraining loss.
发表机构
- Wuhan University（武汉大学）
- The Hong Kong Polytechnic University（香港理工大学）
机构由 AI 辅助整理，请以论文原文为准。
