Research map / Video / Video segmentation and tracking
Online Reasoning Video Object Segmentation
jinyuan Liu, Yang Wang, Zeyu Zhao, Weixin Li, Song Wang, Ruize Han
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 continually-updating segmentation prompt (without affinity fusion) (ablation) | Online ORVOS baseline · ORVOSB | J&F: 52.4% -> 55.1% | Improved |
| add continually-updating segmentation prompt (without affinity fusion) (ablation) | Online ORVOS baseline · ReVOS | J&F: 52.0% -> 53.0% | Improved |
| add affinity-guided adaptive fusion to CUSP (ablation) | Online ORVOS baseline · ORVOSB | J&F: 55.1% -> 58.1% | Improved |
| add affinity-guided adaptive fusion to CUSP (ablation) | Online ORVOS baseline · ReVOS | J&F: 53.0% -> 53.7% | Improved |
| add token reservoir with uniform retention (ablation) | Online ORVOS baseline · ORVOSB | J&F: 58.1% -> 60.9% | Improved |
| add token reservoir with uniform retention (ablation) | Online ORVOS baseline · ReVOS | J&F: 53.7% -> 54.1% | Improved |
| replace uniform retention with dense-to-sparse retention (ablation) | Online ORVOS baseline · ORVOSB | J&F: 60.9% -> 61.3% | Improved |
| replace uniform retention with dense-to-sparse retention (ablation) | Online ORVOS baseline · ReVOS | J&F: 54.1% -> 54.3% | Improved |
| use online causal inference method versus offline global-video method VRS-HQ | Ours · ReVOS | J&F: 59.1% (VRS-HQ) vs 54.3% (Ours) | Hurt |
| compare proposed method against READ on online benchmark | Ours · ORVOSB | J&F: 51.5% -> 61.3% | Improved |
| segment-then-track paradigm (VISA, VRS-HQ) versus frame-specific prompts (LISA, READ) | baselines · ORVOSB | J&F: 33.3% and 43.4% vs 46.6% and 51.5% | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- READ — READ: Reciprocal Attention Discriminator for Image-to-Video Re-Identification (ECCV 2020) · better on 8 of 8 shared results
- GLUS — GLUS: Global-Local Reasoning Unified into A Single Large Language Model for Video Segmentation (CVPR 2025) · better on 1 of 2 shared results
- VISA — VISA: Reasoning Video Object Segmentation via Large Language Model (ECCV 2024) · 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
- Beyond Token Representations: Explicit Visual Object Grounding for Video Reasoning Segmentation (NeurIPS 2026 · similarity 0.90)
- VideoSEG-O3: A Multi-turn Reinforcement Learning Framework for Reasoning Video Object Segmentation (ICML 2026 · similarity 0.90)
- The Devil is in Temporal Token: High Quality Video Reasoning Segmentation (CVPR 2025 · similarity 0.89)
- Fast Reasoning Segmentation for Images and Videos (CVPR 2026 · similarity 0.89)
- One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos (NeurIPS 2024 · similarity 0.89)
- VideoEspresso: A Large-Scale Chain-of-Thought Dataset for Fine-Grained Video Reasoning via Core Frame Selection (CVPR 2025 · 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.
