Research map / Video / Video segmentation and tracking
Per-Clip Video Object 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 |
|---|---|---|---|
| Use per-clip inference with intra-clip refinement, progressive matching, and per-clip training | PCVOS · YouTube-VOS 2019 val | Overall: 82.7 -> 84.6 | Helped |
| Use per-clip inference with the proposed modules | PCVOS · DAVIS 2017 val | J&F: 85.4 -> 86.1 | Helped |
| Use per-clip inference with the proposed modules | PCVOS · DAVIS 2016 val | J&F: 91.6 -> 91.9 | Helped |
| Add progressive memory matching (ablation) | STCN · YouTube-VOS 2019 val, L=15 | Overall: 79.6 -> 81.7 | Helped |
| Add progressive memory matching (ablation) | STCN · YouTube-VOS 2019 val, L=5 | Overall: 82.7 -> 82.7 | No clear effect |
| Add per-clip training (ablation) | STCN with PMM · YouTube-VOS 2019 val, L=5 | Overall: 82.7 -> 83.6 | Helped |
| Add per-clip training (ablation) | STCN with PMM · YouTube-VOS 2019 val, L=15 | Overall: 81.7 -> 82.5 | Helped |
| Add intra-clip refinement (ablation) | PCVOS with PMM and PCT · YouTube-VOS 2019 val, L=5 | Overall: 83.6 -> 84.6 | Helped |
| Add intra-clip refinement (ablation) | PCVOS with PMM and PCT · YouTube-VOS 2019 val, L=15 | Overall: 82.5 -> 83.6 | Helped |
| Train frame-by-frame instead of clip-wise (ablation) | PCVOS · YouTube-VOS 2019 val, L=5 | Overall: 84.6 -> 83.5 | Hurt |
| Remove clip-level supervision (ablation) | PCVOS · YouTube-VOS 2019 val, L=5 | Overall: 84.6 -> 83.7 | Hurt |
| Remove the spatial locality constraint from intra-clip refinement (ablation) | PCVOS · YouTube-VOS 2019 val, L=5 | Overall: 84.6 -> 80.8 | Hurt |
Closest papers by meaning
- Classifying, Segmenting, and Tracking Object Instances in Video with Mask Propagation (CVPR 2020 · similarity 0.89)
- Fast Video Object Segmentation by Reference-Guided Mask Propagation (CVPR 2018 · similarity 0.89)
- Learning Position and Target Consistency for Memory-Based Video Object Segmentation (CVPR 2021 · similarity 0.89)
- Learning Dynamic Network Using a Reuse Gate Function in Semi-Supervised Video Object Segmentation (CVPR 2021 · similarity 0.89)
- From ViT Features to Training-free Video Object Segmentation via Streaming-data Mixture Models (NeurIPS 2023 · similarity 0.89)
- Video Instance Segmentation using Inter-Frame Communication Transformers (NeurIPS 2021 · similarity 0.89)
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.