markadivalerio/audio-classifier-project

Project for UT Scalable Machine Learning Fall 2019

audio-classifier-project's Issues

Can you help me to setup your model with my data ?

Greetings,
Hope you have a nice day, i am working on an audio classification project which i have to distinguish between very similar sounds so i thought about your approach of using the multiple futures but didn't know how, so far i used only mel-spectgrom and barely got a 32% error rate with image classification (using fastai)
so i want to try your model with my data if that is possible but, i didn't understand the code that much.

i wish to contact you in order to help me setup the code, here is my discord
(Artist201#5802)
hope to meet you as soon as possible
Thank you.

question about feature extraction

i see in feature extraction after this piece of code

audio_data, sample_rate = librosa.load(file_name) stft = np.abs(librosa.stft(audio_data))

that you some time use stf and some times use audio_data
like

mfcc = np.mean(librosa.feature.mfcc(y=audio_data, sr=sample_rate, n_mfcc=40).T,axis=0)
    features.extend(mfcc) # 40 = 40  

chroma = np.mean(librosa.feature.chroma_stft(S=stft, sr=sample_rate).T,axis=0)
features.extend(chroma) # 12 = 52`

why is that ? what is the difference ? would like you to elaborate more on this. :)

Thank you !