用自监督重打光提升单目深度估计精度，适配真实场景。
ReDepth Anything: Test-Time Depth Refinement via Self-Supervised Re-lighting
- 测试时通过重打光生成新图像，利用扩散模型先验优化深度图。
- 在多个基准上比DA-V2提升显著，达最新最优水平。
- 适合需高精度深度的3D重建、机器人导航等场景。
单目深度估计仍具挑战性，因为如Depth Anything V2（DA-V2）等基础模型在远离训练分布的真实图像上表现不佳。我们提出Re-Depth Anything，一种测试时自监督框架，通过融合基础模型与大规模2D扩散模型的强大先验来弥合领域差距。该方法通过重打光预测的深度图并增强输入图像，直接在输入图像上进行无标签精炼。此重合成方法采用得分蒸馏采样（SDS），以生成式上下文替代传统光度重建，利用形状从明暗（SfS）线索。为防止优化崩溃，框架仅更新中间嵌入和解码器权重，而非直接优化深度张量或微调全模型。在多个基准上，Re-Depth Anything相较于DA-V2显著提升深度准确率与真实感；应用于Depth Anything 3（DA3）之上，达到当前最优结果，展示了基于几何推理的自监督新路径。
原文摘要 · Abstract (English)
Monocular depth estimation remains challenging, as foundation models such as Depth Anything V2 (DA-V2) struggle with real-world images that are far from the training distribution. We introduce Re-Depth Anything, a test-time self-supervision framework that bridges this domain gap by fusing foundation models with the powerful priors of large-scale 2D diffusion models. Our method performs label-free refinement directly on the input image by re-lighting the predicted depth map and augmenting the input. This re-synthesis method replaces classical photometric reconstruction by leveraging shape from shading (SfS) cues in a new, generative context with Score Distillation Sampling (SDS). To prevent optimization collapse, our framework updates only intermediate embeddings and the decoder's weights, rather than optimizing the depth tensor directly or fine-tuning the full model. Across diverse benchmarks, Re-Depth Anything yields substantial gains in depth accuracy and realism over DA-V2, and applied on top of Depth Anything 3 (DA3) achieves state-of-the-art results, showcasing new avenues for self-supervision by geometric reasoning.
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。