Research map / 3D vision / Depth and stereo
UniDAC: Universal Metric Depth Estimation for Any Camera
Girish Chandar Ganesan, Yuliang Guo, Liu Ren, Xiaoming Liu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add relative/scale decoupling with scale map instead of direct metric depth (ablation) | UniDAC · KITTI-360 | delta1: 0.563 -> 0.622 | Improved |
| add relative/scale decoupling with scale map instead of direct metric depth (ablation) | UniDAC · ScanNet++ | delta1: 0.782 -> 0.792 | Improved |
| use scale map S instead of single scalar scale (ablation) | UniDAC · KITTI-360 | delta1: 0.601 -> 0.622 | Improved |
| use scale map S instead of single scalar scale (ablation) | UniDAC · ScanNet++ | delta1: 0.773 -> 0.792 | Improved |
| single scalar scale vs direct metric depth (ablation) | UniDAC · ScanNet++ | delta1: 0.782 -> 0.773 | Hurt |
| replace 2D RoPE with RoPE-phi (ablation) | UniDAC · ScanNet++ | delta1: 0.750 -> 0.792 | Improved |
| replace 2D RoPE with RoPE-phi (ablation) | UniDAC · KITTI-360 | delta1: 0.592 -> 0.622 | Improved |
| use depth-guided shift map T instead of scalar shift (ablation) | UniDAC · KITTI-360 | delta1: 0.622 -> 0.630 | Improved |
| initialize encoder with DINOv3 instead of DINOv2 (ablation) | UniDAC · KITTI-360 | delta1: 0.445 -> 0.622 | Improved |
| initialize encoder with DINOv3 instead of DINOv2 (ablation) | DAC_U · KITTI-360 | delta1: 0.334 -> 0.428 | Improved |
| add A2D2 to training data (ablation) | UniDAC (GT intrinsics) · KITTI-360 | delta1: 0.757 -> 0.836 | Improved |
| use predicted (AnyCalib) instead of ground-truth intrinsics (ablation) | UniDAC · KITTI-360 | delta1: 0.836 -> 0.815 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- UniK3D — UniK3D: Universal Camera Monocular 3D Estimation (CVPR 2025) · better on 3 of 4 shared results
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
- UniDepth: Universal Monocular Metric Depth Estimation (CVPR 2024 · similarity 0.93)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.91)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.90)
- Metric from Human: Zero-shot Monocular Metric Depth Estimation via Test-time Adaptation (NeurIPS 2024 · similarity 0.90)
- Depth Any Camera: Zero-Shot Metric Depth Estimation from Any Camera (CVPR 2025 · similarity 0.89)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 2024 · similarity 0.89)
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.
