Research map / Video / Video segmentation and tracking
VIRST: Video-Instructed Reasoning Assistant for SpatioTemporal Segmentation
Jihwan Hong, Jaeyoung Do
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 Initial and Second ST-Fusion together (vs. removing Initial) (ablation) | VIRST · MeViS valid | J&F: 59.7 -> 62.9 | Improved |
| use Second ST-Fusion with cross-attention instead of an MLP (ablation) | VIRST · MeViS valid | J&F: 59.4 -> 62.9 | Improved |
| use dynamic anchor updating instead of first-frame anchor (ablation) | VIRST · MeViS valid | J&F: 57.9 -> 62.9 | Improved |
| use dynamic anchor updating instead of CLIP-guided selection (ablation) | VIRST · MeViS valid | J&F: 59.3 -> 62.9 | Improved |
| update anchors over time instead of fixed uniform-3 anchors (ablation) | VIRST · MeViS valid | J&F: 61.9 -> 62.9 | Improved |
| increase number of anchor frames from 2 to 8 (ablation) | VIRST · MeViS valid | J&F: 62.5 -> 63.2 | Improved |
| increase anchor frames from 3 to 4 (ablation) | VIRST · MeViS valid | J&F: 62.9 -> 62.9 | No clear effect |
| VIRST vs prior best on image reasoning segmentation | VIRST · ReasonSeg | cIoU: 65.2 -> 65.0 | No clear effect |
| VIRST vs prior best on RefCOCO testB | VIRST · RefCOCO testB | cIoU: 84.0 -> 81.2 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 4 on most shared results.
- MPG-SAM2 — MPG-SAM 2: Adapting SAM 2 with Mask Priors and Global Context for Referring Video Object Segmentation (ICCV 2025) · better on 2 of 2 shared results
- InstructSeg — InstructSeg: Unifying Instructed Visual Segmentation with Multi-modal Large Language Models (ICCV 2025) · better on 1 of 2 shared results
- SAMWISE — SAMWISE: Infusing Wisdom in SAM2 for Text-Driven Video Segmentation (CVPR 2025) · better on 1 of 1 shared result
- VISA-13B — VISA: Reasoning Video Object Segmentation via Large Language Model (ECCV 2024) · better on 1 of 1 shared result
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
- Reinforcing Video Object Segmentation to Think before it Segments (CVPR 2026 · similarity 0.92)
- Semantic and Sequential Alignment for Referring Video Object Segmentation (CVPR 2025 · similarity 0.89)
- The Devil is in Temporal Token: High Quality Video Reasoning Segmentation (CVPR 2025 · similarity 0.89)
- Motion-Grounded Video Reasoning: Understanding and Perceiving Motion at Pixel Level (CVPR 2025 · similarity 0.89)
- One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos (NeurIPS 2024 · similarity 0.88)
- Unleashing the Temporal-Spatial Reasoning Capacity of GPT for Training-Free Audio and Language Referenced Video Object Segmentation (AAAI 2025 · similarity 0.88)
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.
