Can-Detect-A-Beer-Can-Classification-Model

Quickstart Guide for Running The Model

Dataset Download

• Dataset: The raw image dataset and the cropped dataset used by the model can be found on Mega.nz.
• Required File: Download and unarchive Beer_Images_Cropped.zip. Ensure the resulting Beer_Images_Cropped folder is in the same directory as the notebooks.

Running the Model

The model is designed to be user-friendly with minimal configurations:

1. Clone the Repository: Clone the repository to your local machine.
2. Jupyter Notebooks:
   • For brew type classification: Run Multiclass_BrewType_Classification.ipynb.
   • For taste labelling: Run Multilabel_Taste_Classification.ipynb.

Additional Information

• beer_glossary.csv: Contains information from the raw dataset, including bounding boxes, beer can details, and image filenames from Beer_Images.zip.
• UQ_Beer_Repo.xlsx: Provides an overview of each unique beer product in the dataset.