Research map / Video / Video segmentation and tracking
Frame-to-Frame Aggregation of Active Regions in Web Videos for Weakly Supervised Semantic Segmentation
Jungbeom Lee, Eunji Kim, Sungmin Lee, Jangho Lee, Sungroh Yoon
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 aggregated web video data (ablation) | SEC · VOC 2012 val | mIoU: 50.7 -> 59.5 | Improved |
| add aggregated web video data (ablation) | DSRG · VOC 2012 val | mIoU: 59.0 -> 62.1 | Improved |
| add aggregated web video data (ablation) | FickleNet · VOC 2012 val | mIoU: 61.2 -> 63.2 | Improved |
| add domain adaptation fine-tuning (ablation) | SEC with video · VOC 2012 val | mIoU: 59.5 -> 61.1 | Improved |
| add domain adaptation fine-tuning (ablation) | DSRG with video · VOC 2012 val | mIoU: 62.1 -> 62.9 | Improved |
| add domain adaptation fine-tuning (ablation) | FickleNet with video · VOC 2012 val | mIoU: 63.2 -> 63.9 | Improved |
| add CRF post-processing (ablation) | Ours VGG-16 · VOC 2012 val | mIoU: 62.8 -> 63.9 | Improved |
| aggregate a single frame (K=1) instead of no web video (K=0) (ablation) | FickleNet · VOC 2012 val | mIoU | No clear effect |
| increase aggregated frames K up to 5 (ablation) | FickleNet · VOC 2012 val | mIoU | Improved |
| increase aggregated frames K above 5 (ablation) | FickleNet · VOC 2012 val | mIoU | No clear effect |
| increase web video samples beyond 15,000 (ablation) | FickleNet · VOC 2012 val | mIoU | No clear effect |
| add web video data to FickleNet on VGG-16 (15,000 samples) (ablation) | FickleNet · VOC 2012 val | mIoU: 61.2 -> 63.9 | Improved |
Closest papers by meaning
- Weakly Supervised Semantic Segmentation Using Web-Crawled Videos (CVPR 2017 · similarity 0.91)
- Webly Supervised Semantic Segmentation (CVPR 2017 · similarity 0.89)
- Weakly Supervised Instance Segmentation for Videos With Temporal Mask Consistency (CVPR 2021 · similarity 0.89)
- Efficient Semantic Video Segmentation with Per-frame Inference (ECCV 2020 · similarity 0.88)
- Weakly-Supervised Semantic Segmentation Using Motion Cues (ECCV 2016 · similarity 0.88)
- Semantic Video Segmentation by Gated Recurrent Flow Propagation (CVPR 2018 · 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.
