BoedayaML

How to use this ML model:

1. First, place your dataset in 'Dataset Preparation/Dataset/' directory.

2. Run 'audio_length_adjustment.ipynb' inside Dataset Preparation folder and wait for it to finish.

3. Run 'audio_data_augmentation.ipynb' inside Dataset Preparation folder and also wait for it to finish.

4. Run 'dataset-labelling.ipynb' inside Dataset Preparation folder and once again, wait for it to finish.

5. Locate 'dataset.json', it should be located inside the same folder as 'dataset-labelling.ipynb' file, then copy it to 'Model' directory.

6. Run 'model.ipynb' inside Model folder and wait for the training to be finished, an h5 file should be created automatically.

7. Your own ML model is done and saved in h5 format. If you want to run prediction on the model, run 'prediction.ipynb' and pick the audio file to predict.

Library used:

1. Librosa 0.8.0

2. SoundFile 0.10.3.post1

3. Numpy 1.19.5

4. Matplotlib 3.4.1

5. Tensorflow 2.5.0

6. Scikit Learn 0.24.2