AnyCam-MCT — Multi-Frame Calibration Transformer

Final checkpoint of the master's thesis "Learning Camera Geometry from Unlabeled Real-World Dynamic Video" (TU Munich, 2026): AnyCam (CVPR 2025) extended with the Multi-Frame Calibration Transformer (MCT), a ~25M-parameter cross-frame attention module that aggregates AnyCalib features over multiple frames for video calibration, coupled to AnyCam's pose head via a focal embedding. Trained fully self-supervised on ~82k frames of in-the-wild video; this checkpoint includes the August 2026 correction pass (evaluation bugs fixed, calibration branch retrained with proper input normalization, checkpoint selected by validation loss).

Code, benchmarks, and reproduction instructions: https://github.com/kalman17/anycam-extension

Headline results (honest protocol, raw rows in the GitHub repo)

Measurement	This model	Best competitor measured
Focal error, native wide KITTI frames (single pass)	3.99 %	15.6 % (VGGT-1B, own preprocessing)
Focal error, KITTI windows	7.2 % (beats AnyCalib on 95 % of windows)	10.4 % (AnyCalib)
Multi-frame gain, KITTI (8 frames)	6.0 %	9.1 % (per-frame averaging)
Rotation error, Sintel (median)	0.40°	0.50° (AnyCam)

Known limitations (stated in the repo README): translation direction on TUM-RGBD is worse than AnyCam; large supervised models (Depth Anything 3) lead absolute pose accuracy; full-trajectory drift vs long-context AnyCam inference.

Files

• mct_final.pt — merged inference checkpoint (pose predictor + MCT calibration branch; load with experiments/benchmark_phase_c_checkpoints.py:create_inference_model
  • load_phase_c_checkpoint from the GitHub repo, input_normalization=True).

License / attribution

Apache-2.0. Contains weights derived from AnyCam (MIT, Wimbauer et al., CVPR 2025) and AnyCalib (Apache-2.0, Tirado-Garín et al., ICCV 2025); DINOv2 backbones (Apache-2.0, Meta). Please cite those works alongside the thesis.