DiFA：擴散模型的推理時前向過程對齊

摘要

当前扩散模型的主流推理框架从根本上将生成过程视为一个数值积分问题。该视角将模型视为精确估计器，忽略了去噪过程中固有的统计不确定性。在本工作中，我们提出前向过程对齐的扩散预测（DiFA），这是一种无需训练的框架，将推理阶段的数据预测精炼重构为序列状态估计问题。与单纯将历史输出用于数值积分不同，DiFA将沿逆推轨迹的迭代数据预测视为相关观测，以建立前向对齐的时间一致性。受卡尔曼滤波启发，该一致性机制根据结构一致性和噪声水平兼容性聚合历史预测。为抑制时间一致性带来的过度平滑倾向，我们引入偏差引导机制，自适应保留残差细节。在CIFAR-10和ImageNet上的实验表明，DiFA在所有评估指标（包括FID、IS和FD-DINOv2）上均取得显著提升，证明将推理过程与前向统计结构对齐能大幅增强生成保真度。

English

The prevailing inference framework for diffusion models formulates generation fundamentally as a problem of numerical integration. This perspective casts the model as an exact estimator, neglecting the inherent statistical uncertainty of the denoising process. In this work, we propose Forward-Process Aligned Diffusion prediction (DiFA), a training-free framework that reframes inference-time data prediction refinement as a sequential state estimation problem. Rather than reusing past outputs solely for numerical integration, DiFA treats iterative data predictions along the reverse trajectory as correlated observations to build a forward-aligned temporal consensus. Inspired by Kalman filtering, this consensus aggregates historical predictions according to structural consistency and noise-level compatibility. To counteract the over-smoothing tendency of temporal consensus, we introduce a deviation guidance mechanism to adaptively preserve residual details. Empirically, DiFA yields significant improvements on CIFAR-10 and ImageNet across the evaluated metrics, including FID, IS, and FD-DINOv2, demonstrating that aligning inference with the forward statistical structure substantially improves generative fidelity.