This project is about detecting defects on steel surface using Unet. The dataset used for this project is the NEU-DET database.
A collection of public datasets for surface metal defects detection task. Source, and roboflow (ready to use) links are included.
The first to adopt the YOLOF model on a metal defect NEU-DET dataset. Using several image preprocessing methods, including Edge Detection, Denoising, Sharpening, and SR, were experimented on the metal defect NEU-DET dataset. In the proposed optimized batch size, backbone, soft efficient non-maximum sup-pression (Soft-NMS), and threshold of YOLO.
Open-source YOLOv8 steel surface defect detection project with training, inference, GUI, and Apple Silicon support.
SA-MoE-DefectDet: steel surface defect detection with morphology-aware sparse expert fusion on NEU-DET.
A YOLO detector fine-tuned on steel surface defects, exported to ONNX and served on CPU with a review screen for the quality inspector.
Industrial surface defect detection - YOLOv5 end-to-end NEU-DET (traditional CV + ONNX deploy)
工业钢板缺陷检测：YOLOv8/PyTorch 对照实验、NEU-DET 数据审计、ONNX CPU 部署与 FastAPI 工作台，可在 PyCharm 运行。
YOLOv8 steel surface defect detector on NEU-DET: weights, exact data split, results and limitations
YOLOv9 object detection for six classes of surface defect on hot-rolled steel strip, trained by tran
Steel surface defect detection on NEU-DET — YOLOv5 vs Faster R-CNN, scored with one shared COCO-style evaluator
YOLOv8n钢材表面缺陷检测：NEU-DET数据转换、训练对比、独立测试与ONNX本地推理核验。
Two-stage steel surface defect detection on NEU-DET (Faster R-CNN + ResNet-50)
To associate your repository with the neu-det topic, visit your repo's landing page and select "manage topics."