AI Paper Atlas

Research map / 3D vision / Depth and stereo

DiffusionDepth: Diffusion Denoising Approach for Monocular Depth Estimation

Yiqun Duan, Xianda Guo, Zheng Zhu

ECCV 2024 · 3D vision › Depth and stereo · approach: Self-supervised monocular depth · 125 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
reformulate depth as diffusion denoising vs best prior (URCDC-Depth) on official KITTI splitDiffusionDepth · KITTI official offline splitRMSE: 1.528 -> 1.418Improved
diffusion denoising head vs VA-Depth on NYU-Depth-v2DiffusionDepth · NYU-Depth-v2RMSE: 0.304 -> 0.295Improved
diffusion head vs VA-Depth/URCDC-Depth on KITTI online serverDiffusionDepth · KITTI online serverSILog: 9.84 vs 9.85No clear effect
diffuse refined depth latent (self-diffusion) instead of sparse GT (ablation)DiffusionDepth · KITTI official splitRMSE: 12.3772 -> 1.4523Improved
diffuse refined depth instead of dense GT on indoor data (ablation)DiffusionDepth · NYU-Depth-V2RMSE: 0.3041 -> 0.2983No clear effect
reduce inference steps from 20 to 5 without retraining (ablation)DiffusionDepth · NYU-Depth-V2RMSE: 0.298 -> 1.1750Hurt
train with fewer inference steps (5) instead of 20 (ablation)DiffusionDepth · NYU-Depth-V2RMSE: 0.298 -> 0.4366Hurt
use HAHI neck with Swin instead of FPN (ablation)DiffusionDepth · KITTI official splitRMSE: 1.5569 -> 1.4523Improved
use Swin transformer backbone instead of ResNet-50 (ablation)DiffusionDepth · KITTI official splitRMSE: 1.7124 -> 1.5569Improved
use depth latent at x2 downsampling instead of x4 (ablation)DiffusionDepth · KITTI official splitRMSE: 1.508 -> 1.4523Improved

Later papers that compared with this method

1 later accepted paper reports DiffusionDepth as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy