Research map / Detection and segmentation / Weakly supervised detection and segmentation
Semi-Supervised Video Salient Object Detection Using Pseudo-Labels
Pengxiang Yan, Guanbin Li, Yuan Xie, Zhen Li, Chuan Wang, Tianshui Chen, Liang Lin
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 one pseudo-label every five frames to 20% GT | RCRNet+NER · VOS | max F-measure: 0.849 -> 0.861 | Improved |
| add pseudo-labels (35%) to 5% GT | RCRNet+NER · VOS | max F-measure: 0.821 -> 0.847 | Improved |
| add pseudo-labels (35%) to 5% GT | RCRNet+NER · VOS | S-measure: 0.832 -> 0.861 | Improved |
| increase pseudo-labels from 20% to 80% with 20% GT | RCRNet+NER · VOS | max F-measure: 0.861 -> 0.850 | Hurt |
| use all GT labels instead of 20% | RCRNet+NER · VOS | max F-measure: 0.849 -> 0.849 | No clear effect |
| add first non-local block (ablation) | RCRNet · VOS | max F-measure: 0.846 -> 0.853 | Improved |
| add DB-ConvGRU (ablation) | RCRNet + first non-local block · VOS | max F-measure: 0.853 -> 0.856 | Improved |
| add second non-local block (ablation) | RCRNet + non-local + DB-ConvGRU · VOS | max F-measure: 0.856 -> 0.861 | Improved |
| replace ConvGRU with DB-ConvGRU (ablation) | RCRNet with two non-local blocks · VOS | max F-measure: 0.857 -> 0.861 | Improved |
| add RGB-input pseudo-label generator (ablation) | RCRNet+NER · VOS | max F-measure: 0.821 -> 0.832 | Improved |
| add adjacent GT as generator input (ablation) | FGPLG generator · VOS | max F-measure: 0.832 -> 0.838 | Improved |
| add optical flow and GT warping to generator (ablation) | FGPLG generator · VOS | max F-measure: 0.838 -> 0.847 | Improved |
Closest papers by meaning
- Semi-Supervised Video Salient Object Detection Based on Uncertainty-Guided Pseudo Labels (NeurIPS 2022 · similarity 0.92)
- Weakly Supervised Salient Object Detection Using Image Labels (AAAI 2018 · similarity 0.91)
- Weakly Supervised Video Salient Object Detection (CVPR 2021 · similarity 0.91)
- Weakly-Supervised Salient Object Detection Using Point Supervision (AAAI 2022 · similarity 0.90)
- Structure-Consistent Weakly Supervised Salient Object Detection with Local Saliency Coherence (AAAI 2021 · similarity 0.89)
- Motion Guided Attention for Video Salient Object Detection (ICCV 2019 · 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.
