- 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 resultingBeer_Images_Cropped
folder is in the same directory as the notebooks.
The model is designed to be user-friendly with minimal configurations:
- Clone the Repository: Clone the repository to your local machine.
- Jupyter Notebooks:
- For brew type classification: Run
Multiclass_BrewType_Classification.ipynb
. - For taste labelling: Run
Multilabel_Taste_Classification.ipynb
.
- For brew type classification: Run
beer_glossary.csv
: Contains information from the raw dataset, including bounding boxes, beer can details, and image filenames fromBeer_Images.zip
.UQ_Beer_Repo.xlsx
: Provides an overview of each unique beer product in the dataset.