---
license: apache-2.0
tags:
  - object-detection
  - yolov9
  - litter-detection
  - waste-detection
  - littercam
  - computer-vision
datasets:
  - taco
  - openlittermap
base_model: yolov9-c
---

# 🗑️ LitterCam — YOLOv9-C Waste Detection (6-hr run)

Detects **10 classes of roadside litter** from CCTV/dashcam footage.



## 📦 Files

| File | Description |
|------|-------------|
| `best.pt` | ⭐ **Best mAP checkpoint** |
| `last.pt` | Latest epoch — resume training |
| `best.onnx` | ONNX FP16 for edge deployment |

## 🚀 Quick Inference

```python
from ultralytics import YOLO
model = YOLO("best.pt")
results = model("road.jpg", conf=0.35)
results[0].show()
```

## 🏷️ Classes
`cigarette_butt` · `plastic_bottle` · `drinks_can` · `fast_food_packaging`
`plastic_bag` · `coffee_cup` · `glass_bottle` · `paper_waste`
`food_wrapper` · `general_litter`

## ⚙️ Training
- **Arch:** YOLOv9-C | **Epochs:** 80 | **Img:** 640px | **Batch:** 8
- **GPU:** Kaggle Dual T4 | **Mode:** 6-hr max accuracy

*Last updated: 2026-05-01 09:11 UTC*