Research map / Detection and segmentation / Open-set and few-shot detection
Fast User-Guided Video Object Segmentation by Interaction-And-Propagation Networks
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 Interaction-and-Propagation Networks instead of competing challenge methods | Interactive video object segmentation · DAVIS Challenge 2018 interactive track | AUC: 0.549 -> 0.641 | Helped |
| Use Interaction-and-Propagation Networks instead of competing challenge methods | Interactive video object segmentation · DAVIS Challenge 2018 interactive track | J@60: 0.395 -> 0.647 | Helped |
| Add static-image pre-training to DAVIS video training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | AUC: 0.555 -> 0.621 | Helped |
| Add static-image pre-training to DAVIS video training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | J@60s: 0.589 -> 0.637 | Helped |
| Add GyGo and Youtube-VOS video data to DAVIS video training without static-image pre-training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | AUC: 0.555 -> 0.627 | Helped |
| Add GyGo and Youtube-VOS video data to DAVIS video training without static-image pre-training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | J@60s: 0.589 -> 0.657 | Helped |
| Use static-image pre-training and all listed video datasets instead of DAVIS-only training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | AUC: 0.555 -> 0.691 | Helped |
| Use static-image pre-training and all listed video datasets instead of DAVIS-only training (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | J@60s: 0.589 -> 0.734 | Helped |
| Add multi-round training rather than training with one interaction per iteration (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | Jaccard score | Helped |
| Add the feature aggregation module rather than using no reference features (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | Jaccard score | Helped |
| Aggregate reference features rather than use a simple identity connection (ablation) | Interaction-and-Propagation Networks · DAVIS-2017 val | Jaccard score | Helped |
Closest papers by meaning
- BubbleNets: Learning to Select the Guidance Frame in Video Object Segmentation by Deep Sorting Frames (CVPR 2019 · similarity 0.88)
- Memory Aggregation Networks for Efficient Interactive Video Object Segmentation (CVPR 2020 · similarity 0.88)
- Fast Video Object Segmentation by Reference-Guided Mask Propagation (CVPR 2018 · similarity 0.88)
- Large-Scale Interactive Object Segmentation With Human Annotators (CVPR 2019 · similarity 0.88)
- Learning Video Object Segmentation From Static Images (CVPR 2017 · similarity 0.88)
- Learning Spatial-Semantic Features for Robust Video Object Segmentation (ICLR 2025 · 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.