TurtleBot Detection Model

Model Description

This is a YOLOv5-based object detection model trained to detect TurtleBots in images. The model was trained using the TurtleBot Detection Dataset v1, which contains 1,006 labeled images with YOLO-format bounding box annotations.

Base Model: YOLOv5s
Classes: 1 (TurtleBot)
Input Size: 640x640 pixels
Framework: PyTorch / Ultralytics YOLOv5
License: CC-BY-4.0

Training Details

The model was trained on 50 epochs using the following key hyperparameters:

Optimizer: SGD
Initial Learning Rate (lr0): 0.01
Final Learning Rate Factor (lrf): 0.01
Momentum: 0.937
Weight Decay: 0.0005
Batch Size: 16
Augmentations:
- Flip Left-Right (fliplr): 50%
- HSV Saturation (hsv_s): 0.7
- Scale (scale): 0.5
- Mosaic: Enabled
Loss Parameters:
- IoU Threshold (iou_t): 0.2
- Classification Loss (cls): 0.5

Usage

To load the model and perform inference:

from ultralytics import YOLO

# Load the model
model = YOLO("fhahn/turtlebot-yolov5").load()

# Run inference on an image
results = model.predict("test_image.jpg")
results.show()

Dataset

This model was trained on the TurtleBot Detection Dataset v1, available at: https://huggingface.co/datasets/fhahn/turtlebot-detection-dataset-v1

License

This model is released under CC-BY-4.0. If you use this model, please cite:

@misc{turtlebot-detection-dataset-v1,
  author = {Fabian Hahn},
  title = {TurtleBot Detection Dataset v1},
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/fhahn/turtlebot-detection-dataset-v1}}
}