DepthMaster：驯服扩散模型用于单目深度估计
DepthMaster: Taming Diffusion Models for Monocular Depth Estimation
AI总结：
本文提出单步扩散模型 DepthMaster，通过 Feature Alignment 与 Fourier Enhancement 模块及两阶段训练弥合生成式与判别式特征差距，在多数据集上实现深度估计的最优泛化与细节保留。
摘要：
扩散-去噪范式中的单目深度估计展现出令人瞩目的泛化能力，但存在推理速度较低的问题。近期方法采用单步确定性范式来提高推理效率，同时保持相当的性能。然而，这些方法忽视了生成式特征与判别式特征之间的差距，导致结果并非最优。在本工作中，我们提出 DepthMaster，这是一种单步扩散模型，旨在调整生成式特征以适应判别式深度估计任务。首先，为缓解对生成式特征所引入纹理细节的过拟合，我们提出 Feature Alignment 模块，该模块融合高质量语义特征，以增强去噪网络的表征能力。其次，为解决单步确定性框架中细粒度细节缺失的问题，我们提出 Fourier Enhancement 模块，以自适应地平衡低频结构与高频细节。我们采用两阶段训练策略，以充分发挥这两个模块的潜力。第一阶段，我们借助 Feature Alignment 模块专注于学习全局场景结构；第二阶段，我们利用 Fourier Enhancement 模块提升视觉质量。通过这些努力，我们的模型在泛化性和细节保留方面达到了最先进性能，并在多个数据集上优于其他基于扩散的方法。项目页面位于 https://indu1ge.github.io/DepthMaster_page。
英文摘要：
Monocular depth estimation within the diffusion-denoising paradigm demonstrates impressive generalization ability but suffers from low inference speed. Recent methods adopt a single-step deterministic paradigm to improve inference efficiency while maintaining comparable performance. However, they overlook the gap between generative and discriminative features, leading to suboptimal results. In this work, we propose DepthMaster, a single-step diffusion model designed to adapt generative features for the discriminative depth estimation task. First, to mitigate overfitting to texture details introduced by generative features, we propose a Feature Alignment module, which incorporates high-quality semantic features to enhance the denoising network's representation capability. Second, to address the lack of fine-grained details in the single-step deterministic framework, we propose a Fourier Enhancement module to adaptively balance low-frequency structure and high-frequency details. We adopt a two-stage training strategy to fully leverage the potential of the two modules. In the first stage, we focus on learning the global scene structure with the Feature Alignment module, while in the second stage, we exploit the Fourier Enhancement module to improve the visual quality. Through these efforts, our model achieves state-of-the-art performance in terms of generalization and detail preservation, outperforming other diffusion-based methods across various datasets. Our project page can be found at https://indu1ge.github.io/DepthMaster_page.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
