任意条件下的Depth Anything
Depth Anything at Any Condition
- VCIP, School of Computer Science, Nankai University（VCIP，计算机科学学院，南开大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出DepthAnything-AC基础单目深度估计模型，通过无监督一致性正则化微调和空间距离约束，仅用少量无标签数据即可在恶劣天气、光照变化和传感器失真等复杂条件下实现鲁棒的零样本深度估计。
AI中文摘要：
我们提出Depth Anything at Any Condition（DepthAnything-AC），这是一个能够处理多样环境条件的基础单目深度估计（MDE）模型。以往的基础MDE模型在一般场景中取得了令人印象深刻的性能，但在涉及光照变化、恶劣天气和传感器引起的失真等挑战性条件的复杂开放世界环境中表现不佳。为了克服数据稀缺以及无法从损坏图像中生成高质量伪标签的挑战，我们提出了一种无监督一致性正则化微调范式，该范式仅需要相对少量的无标签数据。此外，我们提出了空间距离约束，以显式地强制模型学习图像块级别的相对关系，从而获得更清晰的语义边界和更准确的细节。实验结果表明，DepthAnything-AC在多种基准上具有零样本能力，包括真实世界恶劣天气基准、合成损坏基准和一般基准。
英文摘要：
We present Depth Anything at Any Condition (DepthAnything-AC), a foundation monocular depth estimation (MDE) model capable of handling diverse environmental conditions. Previous foundation MDE models achieve impressive performance across general scenes but not perform well in complex open-world environments that involve challenging conditions, such as illumination variations, adverse weather, and sensor-induced distortions. To overcome the challenges of data scarcity and the inability of generating high-quality pseudo-labels from corrupted images, we propose an unsupervised consistency regularization finetuning paradigm that requires only a relatively small amount of unlabeled data. Furthermore, we propose the Spatial Distance Constraint to explicitly enforce the model to learn patch-level relative relationships, resulting in clearer semantic boundaries and more accurate details. Experimental results demonstrate the zero-shot capabilities of DepthAnything-AC across diverse benchmarks, including real-world adverse weather benchmarks, synthetic corruption benchmarks, and general benchmarks. Project Page: https://ghost233lism.github.io/depthanything-AC-page Code: https://github.com/HVision-NKU/DepthAnythingAC
