Skip to content

deblinbagchi/audio_classification_ICBHI

Repository files navigation

audio_classification_ICBHI

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:

  1. 102_1b1_Ar_sc_Meditron.wav ------> Original label "Healthy" (Can be found here)
  2. 104_1b1_Ll_sc_Litt3200.wav ------> Original label "COPD" (Can be found here)
  3. 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

About

Audio classification task on ICBHI data

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors