SOBE

SERT's Own Bucket Extractor

What?

SOBE is a neural network based off of the YOLO model to detect the locations of buckets within an image. It was originally used in the 2017 Bunnybots competition.

Dataset

The dataset used to train SOBE is available in the SouthEugeneRoboticsTeam/bucket-dataset repository. Feel free to contribute more images and labels!

Usage

Installing

Before installing SOBE, ensure you have Python installed.

Then, clone SOBE locally:

$ git clone https://github.com/SouthEugeneRoboticsTeam/sobe

Navigate to the freshly cloned SOBE directory, and install the requirements:

$ pip install -r requirements.txt

Training

To train the network, first configure the config.json file for your dataset, then run the following command:

$ python train.py -c config.json

Testing

To test the trained weights on a specific image, use the following command:

$ python predict.py -c config.json -i /path/to/image.jpg -w /path/to/weights.h5