Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformers

Gangwei Xu, Haotong Lin, Hongcheng Luo, Xianqi Wang, Jingfeng Yao, Lianghui Zhu, Yuechuan Pu, Cheng Chi_, Haiyang Sun, BING WANG, Guang Chen, Hangjun Ye, Sida Peng, Xin Yang

Keywords: monocular depth estimation; generative depth estimation models; diffusion transformers; relative depth

TL;DR: We present Pixel-Perfect Depth, a monocular depth estimation model based on pixel-space diffusion transformers.

Abstract: This paper presents **Pixel-Perfect Depth**, a monocular depth estimation model based on pixel-space diffusion generation that produces high-quality, flying-pixel-free point clouds from estimated depth maps. Current generative depth estimation models fine-tune Stable Diffusion and achieve impressive performance. However, they require a VAE to compress depth maps into the latent space, which inevitably introduces flying pixels at edges and details. Our model addresses this challenge by directly performing diffusion generation in the pixel space, avoiding VAE-induced artifacts. To overcome the high complexity associated with pixel-space generation, we introduce two novel designs: 1) **Semantics-Prompted Diffusion Transformers** (**SP-DiT**), which incorporate semantic representations from vision foundation models into DiT to prompt the diffusion process, thereby preserving global semantic consistency while enhancing fine-grained visual details; and 2) **Cascade DiT Design** that progressively increases the number of tokens to further enhance efficiency and accuracy. Our model achieves the best performance among all published generative models across five benchmarks, and significantly outperforms all other models in edge-aware point cloud evaluation. Project page: https://pixel-perfect-depth.github.io/.

Primary Area: Deep learning (e.g., architectures, generative models, optimization for deep networks, foundation models, LLMs)

Submission Number: 2623