Research map / Detection and segmentation / Open-set and few-shot detection
Video Panoptic Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 pixel-level Fuse and object-level Track modules together (ablation) | VPSNet · VIPER | VPQ: 34.9 -> 51.9 | Helped |
| Add pixel-level Fuse to the Track variant (ablation) | VPSNet · VIPER | VPQ: 45.8 -> 51.9 | Helped |
| Add pixel-level Fuse to the Track variant (ablation) | VPSNet · Cityscapes-VPS val | VPQ: 55.9 -> 57.0 | Helped |
| Add pixel-level Fuse to the Track variant (ablation) | VPSNet · Cityscapes-VPS test | VPQ: 56.3 -> 57.4 | Helped |
| Add object-level Track to pixel-level Fuse (ablation) | VPSNet · VIPER | PQ: 53.0 -> 55.4 | Helped |
| Use feature alignment without attention (ablation) | VPSNet · VIPER | PQ: 52.1 -> 52.3 | Helped |
| Use feature attention without alignment (ablation) | VPSNet · VIPER | PQ: 52.1 -> 50.7 | Hurt |
| Replace learned tracking with flow-guided mask IoU matching (ablation) | VPSNet · VIPER | VPQ: 51.9 -> 44.6 | Hurt |
| Replace learned tracking with classification-logit sorting (ablation) | VPSNet · VIPER | VPQ: 51.9 -> 36.5 | Hurt |
| Keep fused features away from the tracking branch (ablation) | VPSNet · VIPER | VPQ: 51.9 -> 50.7 | Hurt |
| Use Fuse with VIPER rather than COCO pretraining (ablation) | VPSNet · Cityscapes val | PQ: 61.6 -> 62.2 | Helped |
| Use Fuse with COCO pretraining instead of the base model (ablation) | VPSNet · Cityscapes val | PQ: 60.6 -> 61.6 | Helped |
Closest papers by meaning
- Slot-VPS: Object-Centric Representation Learning for Video Panoptic Segmentation (CVPR 2022 · similarity 0.92)
- Scene-Centric Unsupervised Video Panoptic Segmentation (CVPR 2026 · similarity 0.90)
- Panoptic, Instance and Semantic Relations: A Relational Context Encoder To Enhance Panoptic Segmentation (CVPR 2022 · similarity 0.90)
- Panoptic Segmentation (CVPR 2019 · similarity 0.90)
- Learning To Associate Every Segment for Video Panoptic Segmentation (CVPR 2021 · similarity 0.90)
- Scene-Centric Unsupervised Panoptic Segmentation (CVPR 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.