Welding Defect Detection Model (YOLOv8)
This is a fine-tuned YOLOv8 model for detecting welding defects such as:
- good weld
- bad weld
- defect
Trained using Ultralytics YOLO on a custom dataset of welding images.
Training Details
- Model Type: YOLOv8
- Dataset: The Welding Defect Dataset - v2
- Classes: 3
- Epochs: 100
- Image Size: 640x640
- Batch Size: 16
Usage
from ultralytics import YOLO
model = YOLO("YOUR_HF_USERNAME/welding-defect-yolov8")
results = model("image.jpg")
results[0].show()
- Downloads last month
- 37
from huggingface_hub import hf_hub_download from ultralytics import YOLO # pick the weights file from this repo's "Files and versions" tab weights = hf_hub_download("avinashhm/welding-defect-yolov8", "<weights>.pt") model = YOLO(weights) source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True)