[论文] Beyond Best-of-$k$: Prior-Consistent DAPS for Ambiguous Nonlinear Inve...

论文概要

研究领域: CV 作者: Merve Kocabas, Benjamin A. Burns, Andreas Geiger, Kent Gauen, Sara Fridovich-Keil 发布时间: 2026-10-22 arXiv: 2610.31592

中文摘要

当非线性逆问题（如相位恢复）存在多个仅靠先验区分的测量一致解时，现有扩散采样器的局部近似可能偏向接近先验支持但本身不受先验支持的模式，而best-of-k评估掩盖了这种失败。PC-DAPS在单次采样运行中维护多个候选解，随退火噪声降低根据先验一致性剪除和替换较弱候选，无需访问真实值。PC-DAPS减少了先验不一致样本，改善了非线性逆问题的样本质量，同时保持了线性逆问题的质量和多样性。

原文摘要

Nonlinear inverse problems such as phase retrieval remain challenging when multiple measurement-consistent solutions are distinguishable only by the prior. Existing diffusion-based samplers rely on local approximations that can favor a measurement-consistent mode close to a prior-supported mode but not itself supported by the prior. We introduce PC-DAPS, which maintains multiple candidate solutions within a single sampling run. As annealing noise decreases, PC-DAPS prunes and replaces weaker candidates according to prior consistency, without any ground-truth access.

#论文 #arXiv #CV #小凯