Research map / Label-efficient and robust learning / Recognition and long-tail
Sewer-ML: A Multi-Label Sewer Defect Classification Dataset and Benchmark
Joakim Bruslund Haurum, Thomas B. Moeslund
Check 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 Xie's binary normal/defect classifier before TResNet-L | TResNet-L on Sewer-ML · Sewer-ML test | F2_CIW: 54.75 -> 55.11 | Improved |
| Add Xie's binary normal/defect classifier before TResNet-L | TResNet-L on Sewer-ML · Sewer-ML test | F1_Normal: 79.88 -> 90.94 | Improved |
| Add Xie's binary first stage (ablation) | Xie's two-stage classifier · Sewer-ML test | F2_CIW: 37.83 -> 48.34 | Improved |
| Add Xie's binary first stage (ablation) | Xie's two-stage classifier · Sewer-ML test | F1_Normal: 0.68 -> 90.62 | Improved |
| Add Chen's binary first stage (ablation) | Chen's two-stage classifier · Sewer-ML test | F2_CIW: 41.74 -> 41.74 | No clear effect |
| Add Myrans's binary first stage (ablation) | Myrans's two-stage classifier · Sewer-ML test | F2_CIW: 19.19 -> 4.11 | Hurt |
| Add Myrans's binary first stage (ablation) | Myrans's two-stage classifier · Sewer-ML test | F1_Normal: 0.00 -> 27.48 | Improved |
| Train the second stage on the full dataset, including normal pipes (ablation) | Xie's two-stage classifier · Sewer-ML test | F2_CIW: 48.34 -> 31.82 | Hurt |
| Train the second stage on the full dataset, including normal pipes (ablation) | Xie's two-stage classifier · Sewer-ML test | F1_Normal: 90.62 -> 87.95 | Hurt |
| Train the second stage on the full dataset, including normal pipes (ablation) | Chen's two-stage classifier · Sewer-ML test | F2_CIW: 41.74 -> 43.14 | Improved |
| Train the second stage on the full dataset, including normal pipes (ablation) | Myrans's two-stage classifier · Sewer-ML test | F2_CIW: 4.11 -> 2.61 | Hurt |
| Train the second stage on the full dataset, including normal pipes (ablation) | Myrans's two-stage classifier · Sewer-ML test | F1_Normal: 27.48 -> 27.48 | No clear effect |
Closest papers by meaning
- Cracks in the Foundation: A Civil Infrastructure Dataset to Challenge Vision Foundation Models (NeurIPS 2026 · similarity 0.81)
- CableInspect-AD: An Expert-Annotated Anomaly Detection Dataset (NeurIPS 2024 · similarity 0.80)
- MVREC: A General Few-shot Defect Classification Model Using Multi-View Region-Context (AAAI 2025 · similarity 0.80)
- MicroFiberDetect: An Application for the Detection of Microfibres in Wastewater Sludge Based on CNNs (AAAI 2025 · similarity 0.79)
- WISNet: Pseudo Label Generation on Unbalanced and Patch Annotated Waste Images (CVPR 2025 · similarity 0.78)
- UrbanWaste: In-the-Bin Dataset for Waste Disposal Inspection with Multi-Granularity Hierarchical Labels (AAAI 2025 · similarity 0.78)
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.
