├── LICENSE
├── README.md
└── assets
    └── fig1_teaser.png


/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2025 Tele-AI
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # Metric-Solver: Sliding Anchored Metric Depth Estimation from a Single Image
 2 | 
 3 | In this paper, we introduce Metric-Solver, a novel sliding anchor-based metric depth estimation method that dynamically adapts to varying scene scales.
 4 | 
 5 | ![teaser](assets/fig1_teaser.png)
 6 | 
 7 | Accurate metric depth estimation is vital for many vision tasks but remains challenging due to varying depth scales across scenes. We propose Metric-Solver, a sliding anchor-based method that adaptively normalizes depth into two components: scaled near-field and tapered far-field. A reference depth anchor dynamically adjusts along the depth axis—smaller anchors enhance near-field precision, while larger ones improve far-field estimation. This unified and adaptive representation handles depths from zero to infinity without manual scale tuning. Experiments show Metric-Solver achieves superior accuracy and cross-dataset generalization compared to existing methods.
 8 | 
 9 | ## [Project page](https://tele-ai.github.io/MetricSolver/) |  [Paper](https://arxiv.org/abs/2504.12103)
10 | 
11 | Code is coming soon.
12 | 
13 | 
14 | ## Citation
15 | 
16 | If you find Metric-Solver useful for your work please cite:
17 | ```
18 | @article{wen2025metricsolver
19 |   author    = {Wen,Tao and Wang,Jiepeng and Chen,Yabo and Xu,Shugong and Zhang,Chi and Li,Xuelong},
20 |   title     = {Metric-Solver: Sliding Anchored Metric Depth Estimation from a Single Image},
21 |   journal   = {arXiv preprint arXiv:2504.12103},
22 |   year      = {2025},
23 | }
24 | ```
25 | 


--------------------------------------------------------------------------------
/assets/fig1_teaser.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Tele-AI/MetricSolver/e3bc997230a9c069db2ba98961ebfcc3b6362005/assets/fig1_teaser.png


--------------------------------------------------------------------------------