Papers › ScaleDepth: Decomposing Metric Depth Estimation into Scale Prediction and Relative...

ScaleDepth: Decomposing Metric Depth Estimation into Scale Prediction and Relative Depth Estimation

11 Jul 2024arXiv:2407.08187archive 2025-07-28

Ruijie Zhu, Chuxin Wang, Ziyang Song, Li Liu, Tianzhu Zhang, Yongdong Zhang

Estimating depth from a single image is a challenging visual task. Compared to relative depth estimation, metric depth estimation attracts more attention due to its practical physical significance and critical applications in real-life scenarios. However, existing metric depth estimation methods are typically trained on specific datasets with similar scenes, facing challenges in generalizing across scenes with significant scale variations. To address this challenge, we propose a novel monocular depth estimation method called ScaleDepth. Our method decomposes metric depth into scene scale and relative depth, and predicts them through a semantic-aware scale prediction (SASP) module and an adaptive relative depth estimation (ARDE) module, respectively. The proposed ScaleDepth enjoys several merits. First, the SASP module can implicitly combine structural and semantic features of the images to predict precise scene scales. Second, the ARDE module can adaptively estimate the relative depth distribution of each image within a normalized depth space. Third, our method achieves metric depth estimation for both indoor and outdoor scenes in a unified framework, without the need for setting the depth range or fine-tuning model. Extensive experiments demonstrate that our method attains state-of-the-art performance across indoor, outdoor, unconstrained, and unseen scenes. Project page: https://ruijiezhu94.github.io/ScaleDepth

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

RuijieZhu94/mmdepth officialpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Depth EstimationMonocular Depth Estimation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Monocular Depth Estimation DDAD ScaleDepth-NK Delta < 1.25 0.871 #2 of 4 Archive leaderboard report
Monocular Depth Estimation DDAD ScaleDepth-NK RMSE 6.097 #2 of 4 Archive leaderboard report
Monocular Depth Estimation DDAD ScaleDepth-NK absolute relative error 0.121 #2 of 4 Archive leaderboard report
Monocular Depth Estimation DIML Outdoor ScaleDepth-NK Delta < 1.25 0.058 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIML Outdoor ScaleDepth-NK RMSE 4.344 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIML Outdoor ScaleDepth-NK absolute relative error 1.007 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Indoor ScaleDepth-NK Delta < 1.25 0.447 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Indoor ScaleDepth-NK RMSE 1.443 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Indoor ScaleDepth-NK absolute relative error 0.355 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Outdoor ScaleDepth-NK Delta < 1.25 0.262 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Outdoor ScaleDepth-NK RMSE 8.632 #1 of 1 Archive leaderboard report
Monocular Depth Estimation DIODE Outdoor ScaleDepth-NK absolute relative error 0.562 #1 of 1 Archive leaderboard report
Monocular Depth Estimation Hypersim ScaleDepth-NK Delta < 1.25 0.413 #1 of 1 Archive leaderboard report
Monocular Depth Estimation Hypersim ScaleDepth-NK RMSE 4.825 #1 of 1 Archive leaderboard report
Monocular Depth Estimation Hypersim ScaleDepth-NK absolute relative error 0.381 #1 of 1 Archive leaderboard report
Monocular Depth Estimation IBims-1 ScaleDepth-NK RMSE 0.59 #3 of 4 Archive leaderboard report
Monocular Depth Estimation IBims-1 ScaleDepth-NK absolute relative error 0.164 #3 of 4 Archive leaderboard report
Monocular Depth Estimation IBims-1 ScaleDepth-NK δ1.25 0.778 #3 of 4 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K Delta < 1.25 0.98 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K Delta < 1.25^2 0.998 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K Delta < 1.25^3 1.000 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K RMSE 1.987 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K RMSE log 0.073 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K Sq Rel 0.136 #13 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split ScaleDepth-K absolute relative error 0.048 #13 of 79 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N Delta < 1.25 0.957 #24 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N Delta < 1.25^2 0.994 #24 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N Delta < 1.25^3 0.999 #24 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N RMSE 0.267 #24 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N absolute relative error 0.074 #24 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 ScaleDepth-N log 10 0.032 #24 of 85 Archive leaderboard report
Monocular Depth Estimation SUN-RGBD ScaleDepth-NK Delta < 1.25 0.866 #2 of 3 Archive leaderboard report
Monocular Depth Estimation SUN-RGBD ScaleDepth-NK RMSE 0.359 #2 of 3 Archive leaderboard report
Monocular Depth Estimation SUN-RGBD ScaleDepth-NK absolute relative error 0.129 #2 of 3 Archive leaderboard report
Monocular Depth Estimation Virtual KITTI 2 ScaleDepth-NK Delta < 1.25 0.834 #1 of 1 Archive leaderboard report
Monocular Depth Estimation Virtual KITTI 2 ScaleDepth-NK RMSE 4.747 #1 of 1 Archive leaderboard report
Monocular Depth Estimation Virtual KITTI 2 ScaleDepth-NK absolute relative error 0.12 #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AttentionSoftmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections