Research map / Video / Video segmentation and tracking
Sliced Wasserstein Bridge for Open-Vocabulary Video Instance Segmentation
Zheyun Qin, Deng Yu, Chuanchen Luo, Zhumin Chen
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 spatio-temporal bridge to baseline (ablation) | SWBridge baseline · LV-VIS val | AP: 8.5 -> 15.0 | Improved |
| add text-visual bridge on top of spatio-temporal bridge (ablation) | SWBridge · LV-VIS val | AP: 15.0 -> 19.9 | Improved |
| add contrastive loss L_VT (ablation) | SWBridge · LV-VIS val | AP: 19.9 -> 20.7 | Improved |
| add spatio-temporal bridge to baseline (ablation) | SWBridge baseline · BURST | APn: 2.3 -> 5.4 | Improved |
| add importance weighting to both bridges (ablation) | SWBridge · LV-VIS val | AP: 15.3 -> 20.7 | Improved |
| add importance weighting in spatio-temporal bridge (ablation) | SWBridge · LV-VIS val | AP: 18.5 -> 20.7 | Improved |
| add importance weighting in text-visual bridge (ablation) | SWBridge · LV-VIS val | AP: 19.6 -> 20.7 | Improved |
| use Sliced Wasserstein instead of Huberized distance (ablation) | SWBridge · LV-VIS val | AP: 18.8 -> 20.7 | Improved |
| use Sliced Wasserstein instead of cosine distance (ablation) | SWBridge · LV-VIS val | AP: 18.6 -> 20.7 | Improved |
| use Sliced Wasserstein instead of Wasserstein distance (ablation) | SWBridge · LV-VIS val | AP: 19.5 -> 20.7 | Improved |
| use Sliced Wasserstein instead of Wasserstein distance (ablation) | SWBridge · BURST | APn: 7.0 -> 8.2 | Improved |
| increase number of random projections L (ablation) | SWBridge · LV-VIS, OVIS, BURST | AP | Mixed |
Closest papers by meaning
- Unified Embedding Alignment for Open-Vocabulary Video Instance Segmentation (ECCV 2024 · similarity 0.89)
- Aligning Instance Brownian Bridge with Texts for Open-Vocabulary Video Instance Segmentation (AAAI 2025 · similarity 0.89)
- OpenVIS: Open-vocabulary Video Instance Segmentation (AAAI 2025 · similarity 0.88)
- Minimizing Labeled, Maximizing Unlabeled: An Image-Driven Approach for Video Instance Segmentation (CVPR 2025 · similarity 0.87)
- Improving Semi-Supervised Semantic Segmentation with Sliced-Wasserstein Feature Alignment and Uniformity (CVPR 2025 · similarity 0.86)
- Open-Vocabulary Segmentation with Semantic-Assisted Calibration (CVPR 2024 · 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.
