提出时间配对一致性，降低流匹配采样方差，提升生成质量与效率。
Temporal Pair Consistency for Variance-Reduced Flow Matching
- 通过成对时间步速度预测耦合，仅在估计器层面减少方差。
- 在CIFAR-10和ImageNet上实现更低FID，相同或更低成本下更优性能。
- 无需修改模型结构或求解器，可无缝集成主流生成流程。
连续时间生成模型（如扩散模型、流匹配、修正流）学习随时间变化的向量场，但通常采用独立处理时间步的目标函数，导致估计方差高且采样效率低。现有方法通过显式平滑惩罚、轨迹正则化或修改概率路径与求解器来缓解。本文提出时间配对一致性（TPC），一种轻量级方差减少原则，通过在同一概率路径上耦合成对时间步的速度预测，在不改变模型架构、概率路径或求解器的前提下，仅在估计器层面实现方差控制。理论分析表明，TPC引入二次轨迹耦合正则项，可严格降低梯度方差，同时保持原始流匹配目标。在流匹配中实例化后，TPC在多个分辨率下的CIFAR-10和ImageNet上均提升样本质量与效率，实现相同或更低计算成本下的更低FID，并可自然扩展至现代SOTA流水线，包括噪声增强训练、基于得分的去噪及修正流。
原文摘要 · Abstract (English)
Continuous-time generative models, such as diffusion models, flow matching, and rectified flow, learn time-dependent vector fields but are typically trained with objectives that treat timesteps independently, leading to high estimator variance and inefficient sampling. Prior approaches mitigate this via explicit smoothness penalties, trajectory regularization, or modified probability paths and solvers. We introduce Temporal Pair Consistency (TPC), a lightweight variance-reduction principle that couples velocity predictions at paired timesteps along the same probability path, operating entirely at the estimator level without modifying the model architecture, probability path, or solver. We provide a theoretical analysis showing that TPC induces a quadratic, trajectory-coupled regularization that provably reduces gradient variance while preserving the underlying flow-matching objective. Instantiated within flow matching, TPC improves sample quality and efficiency across CIFAR-10 and ImageNet at multiple resolutions, achieving lower FID at identical or lower computational cost than prior methods, and extends seamlessly to modern SOTA-style pipelines with noise-augmented training, score-based denoising, and rectified flow.
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。