MemSeg
Unofficial Re-implementation for MemSeg: A semi-supervised method for image surface defect detection using differences and commonalities
Environments
- Docker image: nvcr.io/nvidia/pytorch:20.12-py3
einops==0.5.0
timm==0.5.4
wandb==0.12.17
omegaconf
imgaug==0.4.0
Process
1. Anomaly Simulation Strategy
2. Model Process
Run
python main.py --yaml_config ./configs.yaml DATASET.target capsuleDemo
voila "[demo] model inference.ipynb" --port ${port} --Voila.ip ${ip}
Results
TBD
| target | AUROC-image | AUROC-pixel | AUPRO-pixel |
|---|---|---|---|
| leather | 100 | 97.29 | 96.14 |
| pill | 93.67 | 92.47 | 84.14 |
| carpet | 97.87 | 96.55 | 90.74 |
| hazelnut | 99.79 | 93.92 | 92 |
| tile | 100 | 98.79 | 97.09 |
| cable | 81.22 | 67.08 | 52.64 |
| transistor | 95.04 | 72.34 | 68.8 |
| zipper | 98.74 | 88.33 | 75.87 |
| metal_nut | 99.8 | 75.91 | 86.55 |
| grid | 99.25 | 95.42 | 89.53 |
| bottle | 100 | 95.78 | 90.53 |
| capsule | 85.08 | 88.17 | 75.95 |
| wood | 100 | 94.79 | 88.61 |
| Average | 96.19 | 88.99 | 83.74 |
Citation
@article{DBLP:journals/corr/abs-2205-00908,
author = {Minghui Yang and
Peng Wu and
Jing Liu and
Hui Feng},
title = {MemSeg: {A} semi-supervised method for image surface defect detection
using differences and commonalities},
journal = {CoRR},
volume = {abs/2205.00908},
year = {2022},
url = {https://doi.org/10.48550/arXiv.2205.00908},
doi = {10.48550/arXiv.2205.00908},
eprinttype = {arXiv},
eprint = {2205.00908},
timestamp = {Tue, 03 May 2022 15:52:06 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2205-00908.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}