Research map / Video / Video segmentation and tracking
A Deep Moving-Camera Background Model
Guy Erez, Ron Shapira Weber, Oren Freifeld
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 |
|---|---|---|---|
| add conditional autoencoder (CAE) on top of robust mean (ablation) | DeepMCBM Hom · dog-gooses (moving-camera background) | AUC: .955 -> .984 | Improved |
| add CAE (ablation) | DeepMCBM Hom · flamingo (moving-camera background) | AUC: .961 -> .980 | Improved |
| add CAE (ablation) | DeepMCBM Hom · sidewalk (moving-camera background) | AUC: .889 -> .932 | Improved |
| add CAE (ablation) | DeepMCBM Hom · continuousPan (moving-camera background) | AUC: .895 -> .938 | Improved |
| add CAE (ablation) | DeepMCBM Hom · bmx-trees (moving-camera background) | AUC: .916 -> .908 | Hurt |
| add CAE (ablation) | DeepMCBM Hom · boxing-fisheye (moving-camera background) | AUC: .927 -> .898 | Hurt |
| add CAE (ablation) | DeepMCBM Aff · stroller (moving-camera background) | AUC: .877 -> .885 | Improved |
| add CAE (ablation) | DeepMCBM Aff · swing (moving-camera background) | AUC: .880 -> .877 | No clear effect |
| add conditioning of the autoencoder on robust moments (ablation) | CAE · tennis (moving-camera background) | AUC: .701 -> .963 | Improved |
| use homography instead of affine STN with CAE (ablation) | DeepMCBM CAE · stroller (moving-camera background) | AUC: .885 -> .756 | Hurt |
| use homography instead of affine STN with CAE (ablation) | DeepMCBM CAE · sidewalk (moving-camera background) | AUC: .908 -> .932 | Improved |
| vary trimming parameter alpha from 0.3 to between 20% and almost 50% (ablation) | robust moments · all videos (moving-camera background) | AUC | No clear effect |
Closest papers by meaning
- JA-POLS: A Moving-Camera Background Model via Joint Alignment and Partially-Overlapping Local Subspaces (CVPR 2020 · similarity 0.88)
- Leader360V: A Large-scale, Real-world 360 Video Dataset for Multi-task Learning in Diverse Environment (NeurIPS 2025 · similarity 0.84)
- Towards Understanding Camera Motions in Any Video (NeurIPS 2025 · similarity 0.83)
- MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models (CVPR 2025 · similarity 0.83)
- MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos (CVPR 2025 · similarity 0.83)
- Dense Video Object Captioning from Disjoint Supervision (ICLR 2025 · similarity 0.83)
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.
