Research map / 3D vision / Depth and stereo
Deep Geometry-Aware Camera Self-Calibration from Video
Annika Hagemann, Moritz Knorr, Christoph Stiller
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 DroidCalib instead of the second-best self-calibration baseline | monocular video self-calibration · EuRoC raw | median mapping error (pixel): 3.48 -> 0.40 | Improved |
| add self-calibrating bundle adjustment layer to DROID-SLAM under erroneous intrinsics | DROID-SLAM · TartanAir, EuRoC, TUM trajectory estimation | ATE | Improved |
| apply DroidCalib to raw distorted images with unified camera model without retraining | DroidCalib · EuRoC raw | median ATE (m): 0.018 -> 0.014 | No clear effect |
| apply self-calibration to hand-held rolling-shutter TUM sequences | DroidCalib · TUM Freiburg-1 | median mapping error (pixel): 0.42 (EuRoC) -> 3.09 (TUM) | Hurt |
| use deep self-calibration instead of target-based calibration with 20 to 100 images | DroidCalib · EuRoC and TUM | mapping error | No clear effect |
| match images via NetVLAD top-5 instead of sequential matching in COLMAP baseline (ablation) | COLMAP baseline · self-calibration baselines | mapping error | Improved |
Closest papers by meaning
- GeoCalib: Learning Single-image Calibration with Geometric Optimization (ECCV 2024 · similarity 0.88)
- AnyCalib: On-Manifold Learning for Model-Agnostic Single-View Camera Calibration (ICCV 2025 · similarity 0.88)
- A Perceptual Measure for Deep Single Image Camera Calibration (CVPR 2018 · similarity 0.87)
- DiffCalib: Reformulating Monocular Camera Calibration as Diffusion-Based Dense Incident Map Generation (AAAI 2025 · similarity 0.86)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.85)
- LoRA3D: Low-Rank Self-Calibration of 3D Geometric Foundation models (ICLR 2025 · similarity 0.85)
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.
