---
license: agpl-3.0
base_model: ultralytics/yolo11n
dataset: farzadnekouei/trash-type-image-dataset
tags:
  - object-detection
  - yolov11
  - trash-detection
---

# 🧠 Trash Detection Model (YOLOv11n)

This model detects and classifies **trash types (plastic, paper, glass, trash)** using YOLOv11n.  
It was trained on the [Trash Type Image Dataset](https://www.kaggle.com/datasets/farzadnekouei/trash-type-image-dataset) from Kaggle.

---

## 📂 Dataset
- **Source:** [Kaggle Trash Type Image Dataset](https://www.kaggle.com/datasets/farzadnekouei/trash-type-image-dataset)  
- **Classes:** `plastic`, `paper`, `glass`, `trash`  
- **License:** Follows the original dataset's Kaggle terms.  
- The dataset was split into `train` and `val` folders for YOLO format.

---

## ⚙️ Training Details
- **Base model:** `ultralytics/yolo11n.pt`
- **Framework:** [Ultralytics YOLO](https://github.com/ultralytics/ultralytics)
- **Image size:** 320 
- **Epochs:** 30  
- **Batch size:** 2
- **Hardware:** Jetson Orin Nano

---

## 🧩 Usage
```python
from huggingface_hub import hf_hub_download
from ultralytics import YOLO

model_path = hf_hub_download(
    repo_id="Alope/trash-detection-yolo11n",
    filename="best.pt"
)
model = YOLO(model_path)
results = model.predict(source="test.jpg", show=True)