Research map / Video / Video segmentation and tracking
CTVIS: Consistent Training for Online Video Instance Segmentation
Kaining Ying, Qing Zhong, Weian Mao, Zhenhua Wang, Hao Chen, Lin Yuanbo Wu, Yifan Liu, Chengxiang Fan, Yunzhi Zhuge, Chunhua Shen
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 memory bank, momentum-averaged embeddings and noise (full CTVIS) to single-reference baseline (ablation) | CTVIS (ResNet-50) · YTVIS19 | AP: 51.6 -> 55.1 | Improved |
| use momentum-averaged embeddings in the memory bank to build contrastive items (ablation) | CTVIS · YTVIS19 | AP: 52.1 -> 54.2 | Improved |
| add noise to memory bank updates (ablation) | CTVIS · YTVIS19 | AP: 54.2 -> 55.1 | Improved |
| add noise to memory bank updates (ablation) | CTVIS · OVIS | AP: 34.9 -> 35.5 | Improved |
| train with more frames per video (up to 8 or 10) using CTVIS contrastive items (ablation) | CTVIS · OVIS | AP: 31.2 -> 35.5 | Improved |
| train IDOL with multiple reference frames and more training frames (ablation) | IDOL multi-reference (Mask2Former) · OVIS | AP: 31.2 -> 27.3 | Hurt |
| add CTVIS to IDOL with Deformable DETR (ablation) | IDOL (Deformable DETR) · YTVIS19 | AP: 49.5 -> 53.7 | Improved |
| use supplementary background negatives only (no major negatives) (ablation) | CTVIS · YTVIS19 | AP: 55.1 -> 16.5 | Hurt |
| sample supplementary negatives from preceding frame (local) instead of all previous frames (global) (ablation) | CTVIS · OVIS | AP: 33.4 -> 35.5 | Improved |
| add copy&paste and crop augmentation to rotation for pseudo-videos (ablation) | CTVIS pseudo-video training · OVIS* | AP: 27.3 -> 30.5 | Improved |
| add crop to rotation for pseudo-videos (ablation) | CTVIS pseudo-video training · YTVIS21* | AP: 48.5 -> 48.7 | No clear effect |
| train CTVIS on pseudo-videos from 5% of frames instead of MinVIS on 100% | CTVIS · OVIS | AP: 39.4 -> 41.9 | Improved |
Later papers that compared with this method
7 later accepted papers report CTVIS as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- MiNQVIS: Mitigating Noisy Queries for Robust Online Video Instance Segmentation (ECCV 2026 · better on 6 of 6 shared results)
- CAVIS: Context-Aware Video Instance Segmentation (ICCV 2025 · better on 6 of 6 shared results)
- Temporal-aware Query Routing for Real-time Video Instance Segmentation (ICCV 2025 · better on 1 of 1 shared result)
- LOMM: Latest Object Memory Management for Temporally Consistent Video Instance Segmentation (ICCV 2025 · better on 1 of 1 shared result)
- VISAGE: Video Instance Segmentation with Appearance-Guided Enhancement (ECCV 2024 · better on 4 of 5 shared results)
- SyncVIS: Synchronized Video Instance Segmentation (NeurIPS 2024 · better on 4 of 4 shared results)
- Improving Video Segmentation via Dynamic Anchor Queries (ECCV 2024 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MinVIS — MinVIS: A Minimal Video Instance Segmentation Framework without Video-based Training (NeurIPS 2022) · 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
- In Defense of Online Models for Video Instance Segmentation (ECCV 2022 · similarity 0.92)
- MiNQVIS: Mitigating Noisy Queries for Robust Online Video Instance Segmentation (ECCV 2026 · similarity 0.91)
- TCOVIS: Temporally Consistent Online Video Instance Segmentation (ICCV 2023 · similarity 0.90)
- Robust and Consistent Online Video Instance Segmentation via Instance Mask Propagation (AAAI 2025 · similarity 0.89)
- SyncVIS: Synchronized Video Instance Segmentation (NeurIPS 2024 · similarity 0.88)
- Minimizing Labeled, Maximizing Unlabeled: An Image-Driven Approach for Video Instance Segmentation (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.
