Open JamesRick opened 4 years ago
We should try using PCA or NMF on the convolutional layers output for SoundNet that do not have a MaxPooling layer immediately after it.
We should try using PCA or NMF on the convolutional layers output for SoundNet that do not have a MaxPooling layer immediately after it.