Research map / Detection and segmentation / CNN features
Video Object Segmentation Using Global and Instance Embedding Learning
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 instance embedding (ablation) | Video object segmentation model · DAVIS17 val | mean J: 76.4 -> 80.2 | Helped |
| add instance embedding (ablation) | Video object segmentation model · DAVIS17 val | mean F: 81.8 -> 85.3 | Helped |
| add global embedding (ablation) | Video object segmentation model · DAVIS17 val | mean J: 79.0 -> 80.2 | Helped |
| add global embedding (ablation) | Video object segmentation model · DAVIS17 val | mean F: 83.8 -> 85.3 | Helped |
| add temporal attention (ablation) | Video object segmentation model · DAVIS17 val | mean J: 78.9 -> 80.2 | Helped |
| add temporal attention (ablation) | Video object segmentation model · DAVIS17 val | mean F: 84.2 -> 85.3 | Helped |
| add refine decoder (ablation) | Video object segmentation model · DAVIS17 val | mean J: 78.7 -> 80.2 | Helped |
| add refine decoder (ablation) | Video object segmentation model · DAVIS17 val | mean F: 83.7 -> 85.3 | Helped |
| use 6 support images instead of 5 (ablation) | Video object segmentation model · DAVIS17 val | mean J: 80.2 -> 80.0 | No clear effect |
| use only the first support image instead of 5 support images (ablation) | Video object segmentation model · DAVIS17 val | mean J: 80.2 -> 69.8 | Hurt |
| use only the previous support image instead of 5 support images (ablation) | Video object segmentation model · DAVIS17 val | mean J: 80.2 -> 69.4 | Hurt |
| use 2 support images instead of 5 (ablation) | Video object segmentation model · DAVIS17 val | mean J: 80.2 -> 79.9 | No clear effect |
Closest papers by meaning
- Unified Mask Embedding and Correspondence Learning for Self-Supervised Video Segmentation (CVPR 2023 · similarity 0.91)
- Instance Embedding Transfer to Unsupervised Video Object Segmentation (CVPR 2018 · similarity 0.91)
- Look Before You Match: Instance Understanding Matters in Video Object Segmentation (CVPR 2023 · similarity 0.90)
- Dense Unsupervised Learning for Video Segmentation (NeurIPS 2021 · similarity 0.90)
- Point-VOS: Pointing Up Video Object Segmentation (CVPR 2024 · similarity 0.87)
- 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.