Research map / Detection and segmentation / Tracking
Fast and Accurate Online Video Object Segmentation via Tracking Parts
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 |
|---|---|---|---|
| Replace simple part averaging with similarity-based aggregation against first-frame parts (ablation) | FAVOS part segmentation · DAVIS 2016 | J mean: 0.739 -> 0.779 | Helped |
| Combine baseline segmentation with tracking the whole object (ablation) | SFL baseline · DAVIS 2016 | J mean: 0.707 -> 0.696 | Hurt |
| Add part-based tracking, ROI segmentation, and average part aggregation (ablation) | SFL baseline · DAVIS 2016 | J mean: 0.707 -> 0.739 | Helped |
| Add the object tracker to similarity-based part aggregation (ablation) | FAVOS · DAVIS 2016 | J mean: 0.779 -> 0.786 | Helped |
| Add dense CRF refinement (ablation) | FAVOS with similarity aggregation and tracker · DAVIS 2016 | J mean: 0.786 -> 0.824 | Helped |
| Add foreground/background regularization (ablation) | FAVOS · DAVIS 2017 validation | J mean: 0.451 -> 0.462 | Helped |
| Add SPN after foreground/background regularization (ablation) | FAVOS · DAVIS 2017 validation | J mean: 0.462 -> 0.481 | Helped |
| Add refinement after foreground/background regularization and SPN (ablation) | FAVOS · DAVIS 2017 validation | J mean: 0.481 -> 0.546 | Helped |
| Use part-based tracking and ROI segmentation without per-instance finetuning | FAVOS · DAVIS 2017 validation | J mean: 0.442 -> 0.451 | Helped |
Closest papers by meaning
- Fast Online Object Tracking and Segmentation: A Unifying Approach (CVPR 2019 · similarity 0.89)
- Classifying, Segmenting, and Tracking Object Instances in Video with Mask Propagation (CVPR 2020 · similarity 0.89)
- Online Video Object Segmentation via Convolutional Trident Network (CVPR 2017 · similarity 0.89)
- Robust and Consistent Online Video Instance Segmentation via Instance Mask Propagation (AAAI 2025 · similarity 0.88)
- Learning Spatial-Semantic Features for Robust Video Object Segmentation (ICLR 2025 · similarity 0.87)
- SMITE: Segment Me In TimE (ICLR 2025 · 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.