Research map / Detection and segmentation / Weakly supervised detection and segmentation
Label-PEnet: Sequential Label Propagation and Enhancement Networks for Weakly Supervised Instance Segmentation
Weifeng Ge, Sheng Guo, Weilin Huang, Matthew R. Scott
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 proposal calibration and the detection module after multi-label classification (ablation) | Label-PEnet cascade · VOC 2007 test | mAP (%): 26.9 -> 39.1 | Improved |
| add the instance refinement module after the detection module (ablation) | Label-PEnet cascade · VOC 2007 test | mAP (%): 39.1 -> 49.7 | Improved |
| add the final instance segmentation module (ablation) | Label-PEnet cascade · VOC 2007 test | mAP (%): 49.7 -> 51.3 | Improved |
| add forward-backward learning with curriculum on top of cascaded pre-training (ablation) | Label-PEnet · VOC 2007 test | mAP (%): 51.3 -> 53.1 | Improved |
| add forward-backward learning (VOC 2012 intermediate module, classification) (ablation) | Label-PEnet multi-label cls · VOC 2007 test | mAP (%): 26.9 -> 28.0 | Improved |
| add forward-backward learning to the detection module (ablation) | Label-PEnet object det · VOC 2007 test | mAP (%): 39.1 -> 40.3 | Improved |
| progressive cascade on VOC 2012: detection module over classification (ablation) | Label-PEnet cascade · VOC 2012 test | mAP (%): 23.9 -> 36.8 | Improved |
| progressive cascade on VOC 2012: instance segmentation over refinement (ablation) | Label-PEnet cascade · VOC 2012 test | mAP (%): 46.7 -> 49.2 | Improved |
| Label-PEnet versus best prior weakly supervised detector MEFF+FRCNN | MEFF+FRCNN · VOC 2007 test | mAP (%): 51.2 -> 53.1 | Improved |
| Label-PEnet versus MEFF+FRCNN on localization | MEFF+FRCNN · VOC 2007 trainval | CorLoc (%): 67.0 -> 68.2 | Improved |
| Label-PEnet versus MCOF-VGG16 on semantic segmentation | MCOF-VGG16 · VOC 2012 test | mIoU (%): 56.2 -> 57.2 | Improved |
| Label-PEnet (VGG16) versus PRM-VGG16 on instance segmentation | PRM-VGG16 · VOC 2012 val | mAP^r@0.5 (%): 22.0 -> 30.2 | Improved |
Later papers that compared with this method
2 later accepted papers report Label-PEnet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Parallel Detection-and-Segmentation Learning for Weakly Supervised Instance Segmentation (ICCV 2021 · better on 1 of 1 shared result)
- Weakly Supervised Instance Segmentation by Learning Annotation Consistent Instances (ECCV 2020 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Leveraging Instance-, Image- and Dataset-Level Information for Weakly Supervised Instance Segmentation (TPAMI 2020 · similarity 0.90)
- Weakly Supervised Learning of Instance Segmentation With Inter-Pixel Relations (CVPR 2019 · similarity 0.90)
- Multi-Evidence Filtering and Fusion for Multi-Label Classification, Object Detection and Semantic Segmentation Based on Weakly Supervised Learning (CVPR 2018 · similarity 0.89)
- Parallel Detection-and-Segmentation Learning for Weakly Supervised Instance Segmentation (ICCV 2021 · similarity 0.89)
- WISH: Weakly Supervised Instance Segmentation using Heterogeneous Labels (CVPR 2025 · similarity 0.89)
- Beyond Semantic to Instance Segmentation: Weakly-Supervised Instance Segmentation via Semantic Knowledge Transfer and Self-Refinement (CVPR 2022 · 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.
