Support for binaural audio production?
#3
by GokseninYuksel - opened
Hey!
It seems like the model can output two-channel audio, though the spatialization is not the greatest. Has there been stereo/binaural audio during the model's training run?
Kind Regards,
Goksenin Yuksel
The autoencoder and diffusion models are trained on stereo audio and music tracks. We didn't use any explicit binaural/HRTF datasets for training.