通过黎曼残差搜索，实现快速且保真的单步图像编辑。
Bridging the Manifold Gap: Riemannian Residual Line Search for One-Step Image Editing
- 基于能量场的梯度方向修正，结合局部曲率调整更新路径。
- 在 PIE-Bench++ 上 700 样本测试中达到当前最优性能。
- 适合需要快速、高质量图像编辑的应用场景。
单步扩散编辑因无需反演和迭代优化而速度快，但单次更新需兼顾实现目标提示词与保持源图完整性——而固定更新强度无法适应所有编辑类型。本文将此矛盾视为对能量场传输后处理的候选选择问题，而非构建新编辑模型。提出黎曼残差线搜索（Riemannian Residual Line Search）方法：首先估计提示词增量场的局部时间曲率，将修正方向投影回原始一阶能量场传输的更新范数；再从源图到强编辑结果构造小残差路径，保留原一阶输出作为候选之一，最终通过最大化目标提示词与 CLIP 的对齐度选取结果。在涵盖 10 种编辑类型、共 700 样本的 PIE-Bench++ 评估中，本方法在现有单步更新算法中达到最新最佳性能。
原文摘要 · Abstract (English)
One-step diffusion editors are fast because they avoid inversion and iterative optimization, but a single transport update must be aggressive enough to realize the target prompt and conservative enough to preserve the source image--and no fixed update strength satisfies both demands across edit types. We treat this tension as a post-hoc candidate-selection problem on top of energy-field transport rather than as a new editing model. Our proposed method, Riemannian Residual Line Search, first builds a stronger edit by estimating the local time curvature of the prompt-delta field and projecting the corrected direction back onto the update norm of the original first-order energy-field transport estimation. It then forms a small residual path from the source image to this strong edit, retains the original first-order output as one candidate, and picks the final image by maximizing target-prompt CLIP alignment. On a 700-sample PIE-Bench++ evaluation across 10 edit type IDs, our method achieves state-of-the-art (SOTA) performance among current one-step update algorithms.
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。