Research map / Video / Video segmentation and tracking
Unsupervised Video Object Segmentation with Joint Hotspot Tracking
Lu Zhang, Jianming Zhang, Zhe Lin, Radomír Měch, Huchuan Lu, You He
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 multi-level correlation features from levels 2-5 instead of only 4,5 (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 74.5 -> 82.2 | Improved |
| add level 2 and level 1 features beyond levels 3,4,5 (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 78.0 -> 82.2 | Improved |
| add level 1 features beyond levels 2,3,4,5 (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 79.9 -> 82.2 | Improved |
| use weighted correlation instead of concatenating template and search features (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 75.3 -> 82.2 | Improved |
| use weighted pooling of template feature with mask/hotspot instead of original correlation (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 78.3 -> 82.2 | Improved |
| add joint hotspot branch training to mask segmentation (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 81.5 -> 82.2 | Improved |
| add joint mask branch training to hotspot tracking (ablation) | WCS-Net · DAVIS-2016 hotspot tracking | KLD: 2.701 -> 2.592 | Improved |
| add joint mask branch training to hotspot tracking (ablation) | WCS-Net · DAVIS-2016 hotspot tracking | CC: 0.399 -> 0.429 | Improved |
| use ResNet101 backbone instead of EfficientNet (ablation) | WCS-Net · DAVIS-2016 val | J Mean: 82.2 -> 81.7 | No clear effect |
| add dense-CRF post-processing | WCS-Net · Youtube-Objects (average) | J Mean (avg): 70.5 -> 70.9 | Improved |
Closest papers by meaning
- See More, Know More: Unsupervised Video Object Segmentation With Co-Attention Siamese Networks (CVPR 2019 · similarity 0.87)
- Fast Online Object Tracking and Segmentation: A Unifying Approach (CVPR 2019 · similarity 0.87)
- Visual Object Tracking with Discriminative Filters and Siamese Networks: A Survey and Outlook (TPAMI 2022 · similarity 0.86)
- Unsupervised Video Object Segmentation using Motion Saliency-Guided Spatio-Temporal Propagation (ECCV 2018 · similarity 0.86)
- Prototypical Cross-Attention Networks for Multiple Object Tracking and Segmentation (NeurIPS 2021 · similarity 0.85)
- Zero-Shot Video Object Segmentation with Co-Attention Siamese Networks (TPAMI 2020 · 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.
