Research map / Detection and segmentation / Weakly supervised detection and segmentation
A Comprehensive Analysis of Weakly-Supervised Semantic Segmentation in Different Image Domains
Lyndon Chan, Mahdi S. Hosseini, Konstantinos N. Plataniotis
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 |
|---|---|---|---|
| use M7 classification network (larger feature map) instead of VGG16 | HistoSegNet/SEC/DSRG/IRNet · datasets with more ground-truth instances (ADP, DeepGlobe) | mIoU (%): mean difference of 5.22% in mIoU | Improved |
| use VGG16 instead of M7 on datasets with few ground-truth instances | SEC/DSRG/IRNet/HistoSegNet · PASCAL VOC2012 | mIoU (%) | Improved |
| add self-supervised segmentation network (SEC) to Grad-CAM seeds | SEC · ADP | mIoU (%) | Hurt |
| add HistoSegNet (no self-supervision) post-processing to Grad-CAM | HistoSegNet · ADP | mIoU (%) | Improved |
| add self-supervised learning (SEC, DSRG) to Grad-CAM seeds | SEC, DSRG · PASCAL VOC2012 | mIoU (%) | Improved |
| add self-supervised learning (DSRG, IRNet) to Grad-CAM seeds | DSRG, IRNet · DeepGlobe | mIoU (%) | Improved |
| add SEC to Grad-CAM seeds | SEC · DeepGlobe | mIoU (%) | Hurt |
| choose self-supervised method when seed recall is low (<40%) | SEC/DSRG/IRNet · VOC2012, DeepGlobe | mIoU (%) | Improved |
| choose non-self-supervised HistoSegNet when seed recall is high (>=40%) | HistoSegNet · ADP-func, ADP-morph | mIoU (%) | Improved |
| use X1.7 network instead of VGG16 | HistoSegNet · ADP | mIoU (%) | Improved |
Closest papers by meaning
- Semi-supervised Semantic Segmentation via Strong-weak Dual-branch Network (ECCV 2020 · similarity 0.91)
- Universal Weakly Supervised Segmentation by Pixel-to-Segment Contrastive Learning (ICLR 2021 · similarity 0.90)
- Bootstrapping the Performance of Webly Supervised Semantic Segmentation (CVPR 2018 · similarity 0.90)
- Weakly Supervised Semantic Segmentation Using Web-Crawled Videos (CVPR 2017 · similarity 0.90)
- Webly Supervised Semantic Segmentation (CVPR 2017 · similarity 0.89)
- Weakly Supervised Gland Segmentation with Class Semantic Consistency and Purified Labels Filtration (AAAI 2025 · 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.
