Research map / Detection and segmentation / Weakly supervised detection and segmentation
Semantic-Transferable Weakly-Supervised Endoscopic Lesions Segmentation
Jiahua Dong, Yang Cong, Gan Sun, Dongdong Hou
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 adversarial learning to the baseline (ablation) | DeepLab-v3 baseline · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 54.19 -> 59.54 | Improved |
| add pseudo labels to BL+AL (ablation) | BL+AL · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 59.54 -> 62.07 | Improved |
| add semantic representation transfer loss to BL+AL+PL (ablation) | BL+AL+PL · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 62.07 -> 63.96 | Improved |
| add pseudo labels to BL+AL+SRT (ablation) | BL+AL+SRT · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 61.49 -> 63.96 | Improved |
| add adversarial learning to BL+PL+SRT (ablation) | BL+PL+SRT · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 63.58 -> 63.96 | Improved |
| add class balance and super-pixel priors to pseudo labels (ablation) | Ours · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 61.69 -> 63.96 | Improved |
| add class balance in pseudo label generation (ablation) | Ours · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 62.29 -> 63.96 | Improved |
| add super-pixel spatial priors in pseudo label generation (ablation) | Ours · Endoscopic dataset (enteroscopy test, 390 images) | mIoU (%): 63.30 -> 63.96 | Improved |
| add super-pixel spatial priors (ablation) | Ours · SYNTHIA to Cityscapes (16 classes, val) | mIoU (%): 42.3 -> 42.9 | Improved |
| add super-pixel spatial priors (ablation) | Ours · GTA to Cityscapes (19 classes, val) | mIoU (%): 46.7 -> 47.2 | Improved |
| add pseudo labels and SRT to BL+AL (SYNTHIA) (ablation) | BL+AL · SYNTHIA to Cityscapes (16 classes, val) | mIoU (%): 36.3 -> 42.9 | Improved |
| add pseudo labels and SRT to BL+AL (GTA) (ablation) | BL+AL · GTA to Cityscapes (19 classes, val) | mIoU (%): 41.2 -> 47.2 | Improved |
Closest papers by meaning
- What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation (CVPR 2020 · similarity 0.92)
- Collaborative Learning of Semi-Supervised Segmentation and Classification for Medical Images (CVPR 2019 · similarity 0.86)
- Weakly Supervised Gland Segmentation with Class Semantic Consistency and Purified Labels Filtration (AAAI 2025 · similarity 0.86)
- Prototype-Based Image Prompting for Weakly Supervised Histopathological Image Segmentation (CVPR 2025 · similarity 0.85)
- EndoBench: A Comprehensive Evaluation of Multi-Modal Large Language Models for Endoscopy Analysis (NeurIPS 2025 · similarity 0.84)
- Multi-view Masked Contrastive Representation Learning for Endoscopic Video Analysis (NeurIPS 2024 · similarity 0.84)
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.
