yolov8n-mobile-phone

YOLOv8n Mobile Phone Detector

A fine-tuned YOLOv8 Nano model trained on the Datacluster Labs Mobile Phone Image Dataset for object detection of mobile phones in images.

Model Details

• Base Model: YOLOv8n (Nano)
• Task: Object Detection
• Dataset: Datacluster Labs Mobile Phone Image Dataset
• Classes: mobile_phone (1 class)
• Training Date: 2026-02-04

Usage

from ultralytics import YOLO

# Load model from Hugging Face
model = YOLO('huggingface://IndUSV/yolov8n-mobile-phone-detector/pytorch_model.bin')

# Run inference
results = model.predict(source='image.jpg', conf=0.5)

Training Details

• Input Size: 640x640
• Batch Size: 16
• Optimizer: SGD
• Learning Rate: Auto
• Epochs: 50 (with early stopping)
• Device: NVIDIA GPU

Dataset Information

The model was trained on the Datacluster Labs Mobile Phone Image Dataset which contains:

• High-resolution mobile phone images
• Pascal VOC format annotations
• Diverse backgrounds and lighting conditions
• Various phone models and orientations

Dataset splits:

• Training: 80%
• Validation: 10%
• Testing: 10%

Performance

Check results.csv for detailed training metrics and evaluation results.

Citation

If you use this model, please cite:

• YOLOv8: https://github.com/ultralytics/ultralytics
• Dataset: https://www.kaggle.com/datasets/dataclusterlabs/mobile-phone-image-dataset

License

This model is released under the MIT License.