AI Paper Atlas

Research map / 3D vision / Depth and stereo

StableDepth: Scene-Consistent and Scale-Invariant Monocular Depth

Zheng Zhang, Lihe Yang, Tianyu Yang, Chaohui Yu, Xiaoyang Guo, Yixing Lao, Hengshuang Zhao

ICCV 2025 · 3D vision › Depth and stereo · approach: Self-supervised monocular depth · 2 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
add unlabeled pseudo-labeled video data with dual DPT decoders (ablation)labeled-only DAv2-L baseline · BonnAbsRel: 0.071 -> 0.057Improved
add unlabeled pseudo-labeled video data with dual DPT decoders (ablation)labeled-only DAv2-L baseline · ScanNetAbsRel: 0.102 -> 0.097Improved
use single DPT decoder with reciprocal output for video data instead of dual decoders (ablation)StableDepth · BonnAbsRel: 0.057 -> 0.109Hurt
use lightweight CNN head instead of second DPT decoder (ablation)StableDepth · ScanNetdelta1: 0.913 -> 0.829Hurt
add second independent DPT decoder for unlabeled video (ablation)single-DPT variant · Bonndelta1: 0.914 -> 0.976Improved
add pseudo-labeled SA-V videos to Hypersim training (ablation)Hypersim-only model · SintelAbsRel: 0.309 -> 0.326Hurt
add pseudo-labeled SA-V videos to Hypersim training (ablation)Hypersim-only model · KITTIAbsRel: 0.122 -> 0.143Hurt
add pseudo-labeled SA-V videos to Hypersim training (ablation)Hypersim-only model · BonnAbsRel: 0.071 -> 0.057Improved
add VKITTI2, LightWheel, Hypersim-U and MatrixCity data on top of SA-V (ablation)StableDepth · KITTIAbsRel: 0.143 -> 0.114Improved
add video-trained dual-decoder StableDepth training to baseline for temporal consistency (ablation)DAv2-L metric baseline · DAVISMTD: 0.003777 -> 0.003482Improved
use DAv2 instead of DINOv2 initialization (labeled data only) (ablation)labeled-only model · BonnAbsRel: 0.086 -> 0.071Improved
use DAv2 instead of DINOv2 initialization (labeled+unlabeled) (ablation)StableDepth · BonnAbsRel: 0.068 -> 0.057Improved

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