I am working on a project, where I have to label the audio datasets which has thousands of data, each audio data is for one second. I have to label where it is in idle or event happening or noise. I used some tool like Audacity and Labelstudio, I can manually label the audio files which is in .wav but thousands of files it is very time consuming is there any alternative method for it, where I could label data efficiently. Now I was trying to write a python script for labelling but there was no luck, was struck in getting it an output viewing format. Could you please share your experience or any other solution for it as supervised learning .
$\begingroup$
$\endgroup$
1
-
$\begingroup$ For the manual labelling, there are good answers here datascience.stackexchange.com/questions/55566/… $\endgroup$– Jon NordbyCommented Feb 12, 2023 at 14:16
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
You dataset is unsupervised. I would, first, suggest you try to cluster your dataset, and, then, apply any supervised learning methodology.
I leave you this post for further reading.
-
$\begingroup$ Hey Thank you for the input. I have already have the labelled and clustered data. I want to label inside the each audio data, the waveforms. How to keep an threshold and doing it using python script for it or any other method . Please share your view if you have iany input with this. $\endgroup$ Commented Aug 27, 2022 at 9:22