qqueing
Question regarding filter bank
Great job on implementing paper!
Question: why did you use python_speech_features.fbank instead of librosa.feature.melspectrogram ? Both transformations are the same, right?
they are almost same except the fbank discards the IDFT procedure.so it can maintain more source info of the audio.
Thank you for answering, I suspect that you mistaken melspectrogram/filterbank for MFCC. Because as far as I understand you use IDFT only for MFCC. So python_speech_features.fbank should be the same as librosa.feature.melspectrogram right? https://librosa.github.io/librosa/generated/librosa.feature.melspectrogram.html https://librosa.github.io/librosa/generated/librosa.feature.mfcc.html