Research map / Video / Video segmentation and tracking
Learning Video Object Segmentation From Unlabeled Videos
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 multi-granularity video pattern learning to saliency-derived fore-background cues | MuG unsupervised · DAVIS16 val, object-level Z-VOS | mean J: 37.2 -> 58.0 | Helped |
| Add multi-granularity video pattern learning to CAM fore-background cues | MuG weakly supervised · DAVIS16 val, object-level Z-VOS | mean J: 45.3 -> 61.2 | Helped |
| Add frame-granularity fore-background learning (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 40.2 -> 58.0 | Helped |
| Add short-term local coherence learning (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 51.3 -> 58.0 | Helped |
| Add long-term cross-frame semantic matching (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 52.8 -> 58.0 | Helped |
| Add video-level global discriminative representation learning (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 56.4 -> 58.0 | Helped |
| Increase bootstrapping from 1 to 2 training iterations (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 50.8 -> 58.0 | Helped |
| Increase bootstrapping from 2 to 3 iterations (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 58.0 -> 58.0 | No clear effect |
| Increase bootstrapping from 2 to 4 iterations (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 58.0 -> 58.0 | No clear effect |
| Add LaSOT training videos (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 58.0 -> 59.5 | Helped |
| Add CRF post-processing (ablation) | MuG · DAVIS16 val, object-level Z-VOS | mean J: 55.3 -> 58.0 | Helped |
| Use CAM supervision instead of unsupervised training | MuG · DAVIS16 val, O-VOS | mean J: 63.1 -> 65.7 | Helped |
Closest papers by meaning
- Dense Unsupervised Learning for Video Segmentation (NeurIPS 2021 · similarity 0.91)
- Two-Shot Video Object Segmentation (CVPR 2023 · similarity 0.90)
- Point-VOS: Pointing Up Video Object Segmentation (CVPR 2024 · similarity 0.88)
- Multi-Granularity Video Object Segmentation (AAAI 2025 · similarity 0.87)
- Minimizing Labeled, Maximizing Unlabeled: An Image-Driven Approach for Video Instance Segmentation (CVPR 2025 · similarity 0.87)
- Dual Prototype Attention for Unsupervised Video Object Segmentation (CVPR 2024 · similarity 0.86)
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.