Research map / 3D vision / Depth and stereo
Camera model and calibration for SfM
Handles camera calibration, distortion and minimal geometric solvers in large-scale structure from motion.
99 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Depth and stereo, keyword cluster “motion · flow · localization”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
19 papers in 2023–24 and 17 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 6.2% to 4.2%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · High-Quality Depth From Uncalibrated Small Motion Clip (CVPR 2016 · 99 citations)
- 2017 · HSfM: Hybrid Structure-from-Motion (CVPR 2017 · 155 citations)
- 2018 · A Perceptual Measure for Deep Single Image Camera Calibration (CVPR 2018 · 116 citations)
- 2019 · Learning Structure-And-Motion-Aware Rolling Shutter Correction (CVPR 2019 · 52 citations)
- 2020 · Warping Residual Based Image Stitching for Large Parallax (CVPR 2020 · 106 citations)
- 2021 · Towards Rolling Shutter Correction and Deblurring in Dynamic Scenes (CVPR 2021 · 54 citations)
- 2022 · 3D Moments From Near-Duplicate Photos (CVPR 2022 · 18 citations)
- 2023 · Perspective Fields for Single Image Camera Calibration (CVPR 2023 · 54 citations)
- 2024 · Detector-Free Structure from Motion (CVPR 2024 · 59 citations)
- 2025 · Cameras as Relative Positional Encoding (NeurIPS 2025 · 10 citations)
- 2026 · Video Stitching from Multiple Moving Cameras (NeurIPS 2026)
Earliest papers here
- High-Quality Depth From Uncalibrated Small Motion Clip (CVPR 2016 · 99 citations)
- Rolling Shutter Camera Relative Pose: Generalized Epipolar Geometry (CVPR 2016 · 72 citations)
- Theory and Practice of Structure-From-Motion Using Affine Correspondences (CVPR 2016 · 56 citations)
What has been tried
91 papers here report 719 measured changes (a component added, removed or swapped, compared with a baseline): helped 68%, no clear effect 7%, hurt 18%, mixed 7%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| use quantization ratio r = 4 instead of r = 8 | ETH3D · Accuracy (1cm) | 80.24 -> 81.58 | Helped | Detector-Free Structure from Motion (CVPR 2024 · 59 citations) |
| Add feature diffusion rather than identity global encoding | Cambridge Landmarks · position error (cm) | 0.8 -> 0.3 | Helped | GLACE: Global Local Accelerated Coordinate Encoding (CVPR 2024 · 36 citations) |
| Add overlap supervision during training | Map-free test · AUC | 0.74 -> 0.75 | No clear effect | Matching 2D Images in 3D: Metric Relative Pose from Metric Correspondences (CVPR 2024 · 26 citations) |
| add token-level CamRay to PRoPE | RealEstate10K with varying intrinsics · PSNR | 21.42 -> 21.78 | Helped | Cameras as Relative Positional Encoding (NeurIPS 2025 · 10 citations) |
| Use LocoNet pre-training | CSRD-II camera registration · Cam.Pos.Avg (m) | 6.98 -> 0.89 | Helped | From a Bird's Eye View to See: Joint Camera and Subject Registration without the Camera Calibration (CVPR 2024 · 10 citations) |
| add ADPs to the five VPs | SL-MH test · pan error | 19.38 -> 2.20 | Helped | Deep Single Image Camera Calibration by Heatmap Regression to Recover Fisheye Images Under Manhattan World Assumption (CVPR 2024 · 5 citations) |
| add bundle adjustment on K to COLMAPguess | Fountain-P11 · Re | 0.444 -> 0.212 | Helped | Minimal Perspective Autocalibration (CVPR 2024 · 5 citations) |
| add intrinsic low-DOF motion representation | Replica tracking average · ATE-RMSE (cm) | 0.54 -> 0.49 | Helped | Continuous Pose for Monocular Cameras in Neural Implicit Representation (CVPR 2024 · 4 citations) |
| Increase minimum edge score from 0.6 to 0.7 | Alamo · reconstruction time | 28 -> 25 | Helped | Leveraging Camera Triplets for Efficient and Accurate Structure-from-Motion (CVPR 2024 · 4 citations) |
| Vary the number of calibration control points | BabelCalib · projection error | Mixed | Structure-from-Motion with a Non-Parametric Camera Model (CVPR 2025 · 3 citations) | |
| Use hybrid inputs and explicit 3D points instead of relative translations only | KITTI-00 · median camera position error (m) | 5.0 -> 3.0 | Helped | Revisiting Global Translation Estimation with Feature Tracks (CVPR 2024 · 3 citations) |
| add the unprojection loss Lu | MipNeRF-360, LLFF, Tanks & Temples, and Aria Everyday Activities · LPIPS | 0.035 -> 0.021; 0.080 -> 0.069; 0.059 -> 0.056; 0.087 -> 0.076 | Helped | AnyMap: Learning a General Camera Model for Structure-from-Motion with Unknown Distortion in Dynamic Scenes (CVPR 2025 · 2 citations) |
What reviewers will expect
Datasets and baselines named in the 12 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- DAVIS · 2 of 12 papers
Baselines
- COLMAP · 2 of 12 papers
- GLOMAP · 2 of 12 papers
- GeoCalib · 2 of 12 papers
- MASt3R · 2 of 12 papers
- CasualSAM · 2 of 12 papers
- LoFTR · 2 of 12 papers
- Dust3R · 2 of 12 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 36 papers; from OpenAlex and the papers' affiliations).
- ETH Zurich · 4 papers
- Lund University · 3 papers
- University of Washington · 2 papers
- Adobe · 2 papers
Use it
Download BibTeX of 39 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Feed-forward 3D reconstruction (133 papers)
- Learned optical and scene flow (108 papers)
- Scene coordinate regression for localization (112 papers)
- Dense feature matching for correspondence (80 papers)
Elsewhere in the topic
- Self-supervised monocular depth (116 papers)
- Neural light field and illumination imaging (107 papers)
- Cost-volume multi-view stereo (107 papers)
- Non-line-of-sight imaging (93 papers)
- Dense depth prediction and completion (85 papers)
- Foundation models for monocular geometry (81 papers)
- Depth from light field and defocus (60 papers)
- Learned stereo with specialized sensors (53 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).