Research map / Video / Video segmentation and tracking
URVOS: Unified Referring Video Object Segmentation Network with a Large-Scale Benchmark
Seonguk Seo, Joon-Young Lee, Bohyung Han
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 cross-modal attention (ablation) | URVOS · Refer-Youtube-VOS val | J: 38.88 -> 45.27 | Improved |
| add memory attention (ablation) | URVOS · Refer-Youtube-VOS val | J: 39.38 -> 45.27 | Improved |
| add cross-modal attention (ablation) | URVOS · Refer-DAVIS17 val | J: 38.25 -> 47.29 | Improved |
| add memory attention (ablation) | URVOS · Refer-DAVIS17 val | J: 39.43 -> 47.29 | Improved |
| add GRU over multi-frame features to image-based baseline | Baseline · Refer-Youtube-VOS val | J: 33.34 -> 34.79 | Improved |
| use two-stage inference with anchor frame and confidence-selected memory instead of forward propagation (ablation) | URVOS · Refer-Youtube-VOS val | J: 43.13 -> 45.27 | Improved |
| select memory frames by confidence instead of previous T frames (ablation) | URVOS · Refer-Youtube-VOS val | J: 44.58 -> 45.27 | Improved |
| add second-stage mask propagation (ablation) | URVOS · Refer-Youtube-VOS val | J: 41.34 -> 45.27 | Improved |
| iterate second stage up to five times (ablation) | URVOS · Refer-Youtube-VOS val | J: 45.27 -> 45.44 | No clear effect |
| increase training data from 10% to 100% of Refer-Youtube-VOS (ablation) | URVOS · Refer-Youtube-VOS val | J: 30.73 -> 45.27 | Improved |
| pretrain on Refer-Youtube-VOS instead of RefCOCO | URVOS · Refer-DAVIS17 val | J: 41.23 -> 44.29 | Improved |
| fine-tune on Refer-DAVIS17 after pretraining | URVOS · Refer-DAVIS17 val | J: 44.29 -> 47.29 | Improved |
Closest papers by meaning
- Language As Queries for Referring Video Object Segmentation (CVPR 2022 · similarity 0.91)
- LoSh: Long-Short Text Joint Prediction Network for Referring Video Object Segmentation (CVPR 2024 · similarity 0.91)
- ReferEverything: Towards Segmenting Everything We Can Speak of in Videos (ICCV 2025 · similarity 0.90)
- OnlineRefer: A Simple Online Baseline for Referring Video Object Segmentation (ICCV 2023 · similarity 0.90)
- InterRVOS: Interaction-Aware Referring Video Object Segmentation (CVPR 2026 · similarity 0.90)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.87)
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.
