Research map / Video / Video segmentation and tracking
ReferDINO: Referring Video Object Segmentation with Visual Grounding Foundations
Tianming Liang, Kun-Yu Lin, Chaolei Tan, Jianguo Zhang, Wei-Shi Zheng, Jian-Fang Hu
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 to the mask decoder (ablation) | ReferDINO (Swin-T) · MeViS | J&F: 47.6 -> 48.0 | Improved |
| add deformable cross-attention to the mask decoder (ablation) | ReferDINO (Swin-T) · MeViS | J&F: 45.3 -> 48.0 | Improved |
| add memory-augmented tracker to the temporal enhancer (ablation) | ReferDINO (Swin-T) · MeViS | J&F: 47.6 -> 48.0 | Improved |
| add cross-modal temporal decoder (ablation) | ReferDINO (Swin-T) · MeViS | J&F: 45.8 -> 48.0 | Improved |
| prune 50% of queries with confidence-aware strategy (ablation) | ReferDINO · Ref-YouTube-VOS | J&F: 67.6 -> 67.5 | No clear effect |
| prune 50% of queries with confidence-aware strategy (ablation) | ReferDINO · Ref-YouTube-VOS | FPS: 4.9 -> 51.0 | Improved |
| prune 50% of queries with confidence-aware strategy (ablation) | ReferDINO · Ref-YouTube-VOS | FLOPs: 840.3G -> 499.3G | Improved |
| prune 50% of queries with confidence-aware strategy (ablation) | ReferDINO · Ref-YouTube-VOS | GPU memory: 24.0G -> 14.1G | Improved |
| replace confidence-aware pruning with random pruning of 50% queries (ablation) | ReferDINO · Ref-YouTube-VOS | J&F: 67.5 -> 38.3 | Hurt |
| use ReferDINO decoder and enhancer instead of dynamic mask head baseline | G-DINO+DH (Swin-T) · Ref-YouTube-VOS | J&F: 64.2 -> 67.5 | Improved |
| use ReferDINO instead of dynamic mask head baseline | G-DINO-DH (Swin-T) · MeViS | J&F: 45.4 -> 48.0 | Improved |
| use grounding-guided mask decoder instead of dynamic mask head | G-DINO+DH · Ref-YouTube-VOS | training memory: 25.3G -> 14.1G | Improved |
Later papers that compared with this method
5 later accepted papers report ReferDINO as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.
- Long-RVOS: A Comprehensive Benchmark for Long-term Referring Video Object Segmentation (CVPR 2026 · better on 3 of 3 shared results)
- DeRVOS: Decoupling Consistent Trajectory Generation and Multimodal Understanding for Referring Video Object Segmentation (CVPR 2026 · better on 3 of 3 shared results)
- Boosting Text-Driven Video Segmentation via Geometry-Aware Distillation (ECCV 2026 · better on 3 of 3 shared results)
- Deforming Videos to Masks: Flow Matching for Referring Video Segmentation (ICLR 2026 · better on 2 of 2 shared results)
- QST-SAM: Leveraging Cross-modal Instructions for Few-shot Referring Video Object Segmentation (ECCV 2026 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OnlineRefer: A Simple Online Baseline for Referring Video Object Segmentation (ICCV 2023 · similarity 0.92)
- LoSh: Long-Short Text Joint Prediction Network for Referring Video Object Segmentation (CVPR 2024 · similarity 0.91)
- Multi-Level Representation Learning With Semantic Alignment for Referring Video Object Segmentation (CVPR 2022 · similarity 0.91)
- Deforming Videos to Masks: Flow Matching for Referring Video Segmentation (ICLR 2026 · similarity 0.91)
- Semantic and Sequential Alignment for Referring Video Object Segmentation (CVPR 2025 · similarity 0.90)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.89)
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.
