Research map / Video / Video segmentation and tracking
ActionVOS: Actions as Prompts for Video Object Segmentation
Liangyang Ouyang, Ruicong Liu, Yifei Huang, Ryosuke Furuta, Yoichi Sato
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 |
|---|---|---|---|
| use ActionVOS setting with action prompts instead of conventional RVOS | ReferFormer-R101 · VISOR val | n-mIoU: 54.2 -> 19.0 | Improved |
| use ActionVOS instead of RVOS | ReferFormer-R101 · VISOR val | p-mIoU: 67.7 -> 65.4 | Hurt |
| add action prompts to object names in training and inference (ablation) | ActionVOS (RF-R101) · VISOR val | p-mIoU: 56.3 -> 65.4 | Improved |
| train ActionVOS without action prompts (object names only) (ablation) | ActionVOS · VOST val | p-mIoU: 29.3 -> 9.0 | Hurt |
| use ActionVOS with action prompts instead of RVOS on state-changed objects | ReferFormer-R101 · VOST val | p-mIoU: 29.3 -> 32.3 | Improved |
| add classification head with threshold 0.75 (ablation) | ActionVOS (RF-R101) · VISOR val | gIoU: 63.2 -> 70.9 | Improved |
| raise classification threshold from 0.25 to 0.75 (ablation) | ActionVOS (RF-R101) · VISOR val | n-mIoU: 32.7 -> 19.0 | Improved |
| use natural-sentence action prompt instead of comma-joined prompt (ablation) | ActionVOS (RF-R101) · VISOR val | p-cIoU: 70.9 -> 72.4 | Improved |
| use action-guided focal loss instead of standard focal loss (ablation) | ActionVOS (RF-R101) · VISOR val | n-cIoU: 36.2 -> 32.7 | Improved |
| use action-guided focal loss instead of standard focal loss (ablation) | ActionVOS (RF-R101) · VISOR val | p-mIoU: 65.4 -> 65.4 | No clear effect |
| apply labeling module rules as post-process to RVOS (RVOS+S4.2) (ablation) | ReferFormer-R101 · VISOR val | n-mIoU: 19.0 -> 35.2 | Hurt |
Later papers that compared with this method
1 later accepted paper reports ActionVOS as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Robust Egocentric Referring Video Object Segmentation via Dual-Modal Causal Intervention (NeurIPS 2025 · better on 6 of 6 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
- Robust Egocentric Referring Video Object Segmentation via Dual-Modal Causal Intervention (NeurIPS 2025 · similarity 0.89)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 2024 · similarity 0.88)
- Action Scene Graphs for Long-Form Understanding of Egocentric Videos (CVPR 2024 · similarity 0.87)
- Progress-Aware Online Action Segmentation for Egocentric Procedural Task Videos (CVPR 2024 · similarity 0.86)
- OnlineTAS: An Online Baseline for Temporal Action Segmentation (NeurIPS 2024 · similarity 0.86)
- ActFusion: a Unified Diffusion Model for Action Segmentation and Anticipation (NeurIPS 2024 · similarity 0.86)
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.
