Research map / Detection and segmentation / CNN features
Video Object Segmentation with Adaptive Feature Bank and Uncertain-Region Refinement
Read the paperCheck 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 |
|---|---|---|---|
| Use AFB-URR instead of STM | semi-supervised video object segmentation · DAVIS17 validation | J & F mean: 71.6 -> 74.6 | Helped |
| Use AFB-URR instead of STM | semi-supervised video object segmentation · YouTube-VOS18 validation | Overall: 79.4 -> 79.6 | Helped |
| Use adaptive feature-bank management instead of STM's fixed memory | video object segmentation · Long-time Video dataset | J & F mean: 79.3 -> 83.3 | Helped |
| Add URR, including confidence loss and local refinement (ablation) | AFB · DAVIS17 validation | J & F mean: 70.2 -> 74.6 | Helped |
| Add local refinement (ablation) | AFB with confidence loss · DAVIS17 validation | J & F mean: 72.7 -> 74.6 | Helped |
| Add confidence loss (ablation) | AFB with local refinement · DAVIS17 validation | J & F mean: 72.1 -> 74.6 | Helped |
| Train on benchmark videos after pretraining (ablation) | AFB-URR · DAVIS17 validation | J & F mean: 60.9 -> 74.6 | Helped |
| Use adaptive feature-bank management rather than first-frame-only memory (ablation) | AFB-URR · DAVIS17 validation | J & F mean: 66.2 -> 74.6 | Helped |
| Use adaptive feature-bank management rather than latest-frame-only memory (ablation) | AFB-URR · DAVIS17 validation | J & F mean: 68.4 -> 74.6 | Helped |
| Use adaptive feature-bank management rather than first-and-latest-frame memory (ablation) | AFB-URR · DAVIS17 validation | J & F mean: 73.1 -> 74.6 | Helped |
| Use adaptive feature-bank management rather than first-and-latest-5-frames memory (ablation) | AFB-URR · DAVIS17 validation | J & F mean: 71.6 -> 74.6 | Helped |
Closest papers by meaning
- Learning Dynamic Network Using a Reuse Gate Function in Semi-Supervised Video Object Segmentation (CVPR 2021 · similarity 0.88)
- Test-time Training for Matching-based Video Object Segmentation (NeurIPS 2023 · similarity 0.88)
- Learning Position and Target Consistency for Memory-Based Video Object Segmentation (CVPR 2021 · similarity 0.88)
- State-Aware Tracker for Real-Time Video Object Segmentation (CVPR 2020 · similarity 0.88)
- Point-VOS: Pointing Up Video Object Segmentation (CVPR 2024 · similarity 0.86)
- Learning Spatial-Semantic Features for Robust Video Object Segmentation (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.