Audio classification task on ICBHI data
You can directly use the evaluate_from_audio.ipynb file in Google collab and change the path to the model and input files. More info on these below. Otherwise, you can also run the evaluate_from_audio.py file.
Please set up your environment as follows:
pip install -r requirements.txt
In order to test out the model, I am selecting the following audio files from the test set of the ICBHI dataset:
- 102_1b1_Ar_sc_Meditron.wav ------> Original label "Healthy" (Can be found here)
- 104_1b1_Ll_sc_Litt3200.wav ------> Original label "COPD" (Can be found here)
- 109_1b1_Ar_sc_Litt3200.wav ------> Original label "COPD" (Can be found here)
These files can be found at my Google drive. Please download the trained model from my Google Drive as well. The model can be found here
In order to test the model in your own environment, please open the file evaluate_model.py and change the path to the model in line 66
and the path to the audio file in line 63. Afterwards, run the file:
python evaluate_from_audio.py