Research map / 3D vision / Depth and stereo
General Self-Calibration with Varying Intrinsics
Norio Kosaka, Timothy Duff, Tomas Pajdla, Akihiro Sugimoto
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 |
|---|---|---|---|
| use Kruppa DIAC solver instead of linear DAQ | shared-focal self-calibration · Synthetic ff000-f, sigma=1.0 | delta_fg: 0.457 -> 0.511 | No clear effect |
| use Kruppa solver instead of linear DAQ | ff000-f · InFlux real sequences | delta_fg: 0.857 -> 0.725 | Improved |
| use Kruppa solver instead of Lourakis pairwise init | ffuv0-uv · InFlux real sequences | delta_uv: 0.797 -> 0.312 | Improved |
| use Kruppa solver instead of Lourakis pairwise init | ffuv0-uv · InFlux real sequences | delta_fg: 0.805 -> 0.790 | No clear effect |
| replace shared aspect/principal point task ffuv0-uv by varying-focal fguv0-uv | Kruppa solver · Synthetic, sigma=1.0 | delta_fg: 1.737 -> 0.547 | Improved |
| use Kruppa ffuv0-uv instead of COLMAP centre guess | principal point estimation · InFlux edited frames (EIS crop) | delta_uv: 2.483 -> 0.395 | Improved |
| use Kruppa fguv0-uv instead of COLMAP centre guess | principal point estimation · InFlux edited frames (EIS crop) | delta_uv: 2.483 -> 0.665 | Improved |
| use Kruppa ffuv0-uv instead of COLMAP centre guess | focal estimation · InFlux edited frames (EIS crop) | delta_fg: 0.692 -> 0.857 | Hurt |
| increase number of sampled relaxations N from 10 to 1000 (ablation) | Jacobian conditioning · ffuv0-uv synthetic | mean condition number: 2.91e3 -> 3.86e3 | No clear effect |
| add stronger cross-view intrinsic sharing (ablation) | Kruppa minimal systems · algebraic complexity | number of solutions: 372191 -> 40 | Improved |
Closest papers by meaning
- AnyCalib: On-Manifold Learning for Model-Agnostic Single-View Camera Calibration (ICCV 2025 · similarity 0.89)
- Self-Calibrating Neural Radiance Fields (ICCV 2021 · similarity 0.87)
- Tame a Wild Camera: In-the-Wild Monocular Camera Calibration (NeurIPS 2023 · similarity 0.86)
- Minimal Perspective Autocalibration (CVPR 2024 · similarity 0.86)
- InFlux: A Benchmark for Self-Calibration of Dynamic Intrinsics of Video Cameras (NeurIPS 2025 · similarity 0.85)
- From Correspondences to Pose: Non-minimal Certifiably Optimal Relative Pose without Disambiguation (CVPR 2024 · similarity 0.84)
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.
