File size: 2,003 Bytes
ea7789a
bdefd1f
 
 
 
 
 
 
 
 
 
ea7789a
bdefd1f
 
 
 
 
 
ea7789a
bdefd1f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fcec5c1
bdefd1f
fcec5c1
bdefd1f
 
 
fcec5c1
bdefd1f
 
 
 
 
 
 
f3257bf
bdefd1f
f3257bf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
---

language:
  - en
  - ru
tags:
  - computer-vision
  - object-detection
  - yolov8
  - welding
  - ndt
  - defect-detection
license: mit
datasets:
  - synthetic-welding-defects
metrics:
  - precision
  - recall
  - mAP
---


# XVL: X-Ray Vision Lab - Welding Defect Detector

YOLOv8-based model for automated detection of welding defects in X-ray images.

## Model Details
- **Architecture**: YOLOv8n (custom)
- **Input Size**: 512x512
- **Classes**: 5 defect types
- **Training Data**: Synthetic X-ray images (8000+ samples)
- **Validation Data**: Real industrial X-ray scans (200+)

## Performance
| Metric     | Value  | Epoch |
|------------|--------|-------|
| Precision  | 95.6%  | 37    |
| Recall     | 88.9%  | 39    |
| mAP@50     | 93.3%  | 39    |
| mAP@50-95  | 78.8%  | 37    |

## Usage

### With PyTorch
```python

import torch

from models.yolo_custom import load_model



model = load_model(

    weights="best.pt",

    config="config.yaml"

)

With XVL Project

bash

git clone https://github.com/your-username/XVL.git

python scripts/download_weights.py

python src/run.py

Defect Classes

Cracks



Pore clusters



Incomplete fusion



Slag inclusions



Absence of defects



Training Configuration

See config.yaml for full details.



Citation

If you use this model, please reference:



@software{xvl2026,

  title={XVL: X-Ray Vision Lab},

  author={Alex Watchman},

  year={2026},

  url={https://github.com/Passenger1993/XVL}

}



## License

MIT License



Copyright (c) 2024 Your Name



Permission is hereby granted, free of charge, to any person obtaining a copy

of this software and associated documentation files (the "Software"), to deal

in the Software without restriction, including without limitation the rights

to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

copies of the Software, and to permit persons to whom the Software is

furnished to do so, subject to the following conditions: