search

afrancl / BinauralLocalizationCNN

Code to create networks that localize sounds sources in 3D environments
38 stars 12 forks source link

Can you share what HRTF data are you using for training? #3

Open Helmholz opened 2 years ago

Helmholz commented 2 years ago

Hi @afrancl , We are trying to run your model for different HRTF renderings, can you tell us what HRTF you are using for your model training?

afrancl commented 2 years ago

Hi @Helmholz, we use the Garnder-Martin HRTFs with our room simulator during training. The details can be found in the paper in the two subsections of our methods, starting with Training data generation. If you are trying to test new stimuli on the pre-trained model, you can use that standard set of Gardner-Martin HRTFs which you can access here.

I hope this helps!