FlowFM：以流匹配推进暗光光流估计
FlowFM: Advancing Dark Optical Flow Estimation with Flow Matching
把流匹配从生成任务迁移到条件回归任务（暗光光流），用真值约束的全局变换路径加单步去噪替代扩散式迭代采样，并用傅里叶幅度/相位解码器恢复暗光下丢失的运动信息，可作为低光照稠密预测任务的通用加速范式。
原文依据 FlowFM incorporates an implicit Fourier denoising decoder (IFDD) for reliable motion understanding.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论暗光光流估计（DOFE）面临严峻挑战：判别式模型对噪声鲁棒性较差，且难以处理被削弱的运动模式；扩散模型则存在光流场不连续与效率低下的问题。流匹配（FM）虽然高效，但在 DOFE 的条件生成中仍缺乏探索。本文提出 FlowFM，这是首个面向 DOFE 任务的流匹配模型。FlowFM 不采用传统的向量场回归，而是提出在真值光流的约束下估计全局变换路径。它将高斯噪声与真值混合生成带噪光流，随后以初始光流场、代价体积与上下文特征为条件执行单步去噪，从而在精度与效率上达到最优。FlowFM 引入隐式傅里叶去噪解码器（IFDD）以实现可靠的运动理解。借助傅里叶变换，IFDD 利用幅度刻画运动强度、利用相位编码光流场中目标的空间关系，并直接增强幅度以恢复暗光导致的运动信息损失。实验表明，FlowFM 在 FCDN 与 VBOF 基准上显著优于最优方法，刷新了 DOFE 的性能纪录。
英文原文摘要 1291 字符
Dark optical flow estimation (DOFE) faces critical challenges: discriminative models are less robust to noise and struggle with weakened motion patterns, while diffusion models suffer from discontinuous flow fields and low efficiency. Flow matching (FM), though efficient, remains underexplored for conditional generation in DOFE. In this paper, we propose FlowFM, the first flow matching model tailored to DOFE tasks. Instead of conventional vector field regression, FlowFM proposes estimating the global transformation path constrained by the ground truth optical flow. It generates noisy flow by mixing Gaussian noise with ground truth, then performs a one-step denoising process conditioned on the initial flow field, cost volume, and contextual features for optimal accuracy and efficiency. FlowFM incorporates an implicit Fourier denoising decoder (IFDD) for reliable motion understanding. By leveraging the Fourier transform, IFDD uses amplitude to characterize motion intensity and phase to encode target spatial relationships within flow fields, then directly enhances amplitude to restore dark-caused motion information loss. Experiments show that FlowFM significantly outperforms state-of-the-art methods on the FCDN and VBOF benchmarks, setting a new performance record for DOFE.