---
license: mit
tags:
- yolov5
- smoking-detection
- object-detection
- onnx
- pytorch
---

# SmokeGuard Models

YOLOv5 models untuk deteksi aktivitas merokok secara real-time.

## Models

| Model | Format | Size | Description |
|-------|--------|------|-------------|
| `best_checkpoint.pt` | PyTorch | ~40 MB | Model training terbaik |
| `last_checkpoint.pt` | PyTorch | ~40 MB | Checkpoint terakhir |
| `best_checkpoint.onnx` | ONNX (FP32) | ~80 MB | Export ONNX untuk inference |
| `best_checkpoint_int8.onnx` | ONNX (INT8) | ~20 MB | Quantized untuk CPU inference |

## Usage

### PyTorch

```python
import torch

model = torch.hub.load("ultralytics/yolov5", "custom", path="best_checkpoint.pt")
results = model("image.jpg")
results.show()
```

### ONNX Runtime

```python
import onnxruntime as ort
import numpy as np

session = ort.InferenceSession("best_checkpoint_int8.onnx")
input_name = session.get_inputs()[0].name
input_data = np.random.randn(1, 3, 640, 640).astype(np.float32)
output = session.run(None, {input_name: input_data})
```

## Quantization

Untuk quantization ONNX model, lihat notebook `quantisize.ipynb`.

## Citation

```bibtex
@article{IPI4527801,
  title = "IMPLEMENTASI METODE YOLOv5 PADA SISTEM PENDETEKSI ROKOK DI AREA BEBAS ASAP ROKOK",
  journal = "Institut Penelitian Matematika, Komputer, Keperawatan, Pendidikan dan Ekonomi (IPM2KPE)",
  year = "2024",
  author = "Fathoni, Aliffatul Majid; Zuliarso, Eri"
}
```