I would like to utilize the following code to do audio classification within OF. My question is, does anyone know any alternative approaches here? Or a way to train a model by pointing to a directory full of audio files rather than training with recordings like in this example.
https://github.com/ml4a/ml4a-ofx/tree/master/apps/AudioClassifier
Hi, an audio file is at the end the same as a recording. The difference is that for the audio file you get all the samples at once, when loading it, while in the recording (live audio is a better word) you get samples by chunks, a.k.a. buffers. Thus, what you want to do is as simple as loading the audio files, then go through its samples by chunks, and pass to the required function. you are interested in these
in order to load the audio files I recommend you to use ofxAudioFile