Add model card
Browse files
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- pytorch
|
| 5 |
+
- onnx
|
| 6 |
+
- object-detection
|
| 7 |
+
- yolo
|
| 8 |
+
- computer-vision
|
| 9 |
+
- pcb-inspection
|
| 10 |
+
- industrial
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Inspection Engine
|
| 14 |
+
|
| 15 |
+
## Model Description
|
| 16 |
+
Inspection Engine is a YOLO11s model fine-tuned for automated PCB (Printed Circuit Board) defect detection. It identifies manufacturing defects such as missing components, solder bridges, lifted pads, and other anomalies in real-time, enabling automated quality control in electronics manufacturing.
|
| 17 |
+
|
| 18 |
+
## Model Architecture
|
| 19 |
+
- **Base Model**: YOLO11s (Small variant — optimized for speed)
|
| 20 |
+
- **Framework**: PyTorch + ONNX (for cross-platform deployment)
|
| 21 |
+
- **Task**: Object Detection (Defect Localization)
|
| 22 |
+
- **Input**: High-resolution PCB images
|
| 23 |
+
|
| 24 |
+
## Training Details
|
| 25 |
+
- **Dataset**: PCB defect inspection dataset with annotated defect regions
|
| 26 |
+
- **Checkpoint**: `inspection_engine_final3` (best performing run)
|
| 27 |
+
- **Augmentations**: Mosaic, color jitter, random affine transforms
|
| 28 |
+
- **Export**: Exported to ONNX for production deployment
|
| 29 |
+
|
| 30 |
+
## Files
|
| 31 |
+
| File | Description |
|
| 32 |
+
|------|-------------|
|
| 33 |
+
| `best.pt` | Best PyTorch model weights |
|
| 34 |
+
| `best.onnx` | ONNX export for cross-platform/production deployment |
|
| 35 |
+
|
| 36 |
+
## Usage
|
| 37 |
+
|
| 38 |
+
### PyTorch (Ultralytics)
|
| 39 |
+
```python
|
| 40 |
+
from ultralytics import YOLO
|
| 41 |
+
from huggingface_hub import hf_hub_download
|
| 42 |
+
|
| 43 |
+
model_path = hf_hub_download(repo_id='devanshty/Inspection-Engine', filename='best.pt')
|
| 44 |
+
model = YOLO(model_path)
|
| 45 |
+
results = model('pcb_image.jpg')
|
| 46 |
+
results[0].show()
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
### ONNX Runtime
|
| 50 |
+
```python
|
| 51 |
+
import onnxruntime as ort
|
| 52 |
+
import numpy as np
|
| 53 |
+
from huggingface_hub import hf_hub_download
|
| 54 |
+
|
| 55 |
+
onnx_path = hf_hub_download(repo_id='devanshty/Inspection-Engine', filename='best.onnx')
|
| 56 |
+
session = ort.InferenceSession(onnx_path)
|
| 57 |
+
# Prepare input (1, 3, H, W) float32 normalized
|
| 58 |
+
input_name = session.get_inputs()[0].name
|
| 59 |
+
outputs = session.run(None, {input_name: np.zeros((1, 3, 640, 640), dtype=np.float32)})
|
| 60 |
+
```
|
| 61 |
+
|
| 62 |
+
## Download & Use
|
| 63 |
+
|
| 64 |
+
```python
|
| 65 |
+
from huggingface_hub import hf_hub_download
|
| 66 |
+
model_path = hf_hub_download(repo_id='devanshty/Inspection-Engine', filename='best.pt')
|
| 67 |
+
```
|