Any publication?
#5
by sappho192 - opened
Hi, thank you for releasing this model into public.
I'd like to study what changes were made in this 2.0 version compared to the previous model, but I couldn't find any papers related to this.
Is there any way I can find out in detail what has changed?
Thanks in advance.
The biggest diff in the training dataset, plus slightly different augmentations. The training data of 2.0 version includes non-speech audio samples to help the model distinguish between speech and non-speech sounds (such as coughing, laughter, and breathing, etc.)
You can refer to MarbleNet Paper: https://arxiv.org/pdf/2010.13886
sappho192 changed discussion status to closed