改进条件生成中的最优传输，让训练测试更一致。
The Curse of Conditions: Analyzing and Improving Optimal Transport for Conditional Flow-Based Generation
- 在代价矩阵中加入条件加权，实现条件最优传输。
- 在多个数据集上优于现有基线，尤其在计算预算有限时。
- 适用于离散与连续条件，代码开源可用。
小批量最优传输可使无条件流匹配的路径更平滑，从而减少推理时的积分步数和数值求解复杂度。但在条件生成场景中，标准最优传输会忽略条件信息，导致训练时使用有偏先验分布，而测试时却采样自无偏先验，造成训练与测试不一致，性能下降。为此，我们提出条件最优传输 C^2OT，通过在代价矩阵中引入条件加权项来修正分配。实验表明，该方法在 8gaussians-to-moons、CIFAR-10、ImageNet-32x32 与 ImageNet-256x256 上均有效，且在不同函数评估预算下整体表现优于现有基线。代码已公开。
原文摘要 · Abstract (English)
Minibatch optimal transport coupling straightens paths in unconditional flow matching. This leads to computationally less demanding inference as fewer integration steps and less complex numerical solvers can be employed when numerically solving an ordinary differential equation at test time. However, in the conditional setting, minibatch optimal transport falls short. This is because the default optimal transport mapping disregards conditions, resulting in a conditionally skewed prior distribution during training. In contrast, at test time, we have no access to the skewed prior, and instead sample from the full, unbiased prior distribution. This gap between training and testing leads to a subpar performance. To bridge this gap, we propose conditional optimal transport C^2OT that adds a conditional weighting term in the cost matrix when computing the optimal transport assignment. Experiments demonstrate that this simple fix works with both discrete and continuous conditions in 8gaussians-to-moons, CIFAR-10, ImageNet-32x32, and ImageNet-256x256. Our method performs better overall compared to the existing baselines across different function evaluation budgets. Code is available at https://hkchengrex.github.io/C2OT
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。