4 ms·
What do you think is the best neural network currently for processing and possibly generating 44.1 Khz music audio data? If we're stuck with downsampling to 16
by bravura 7y ago
What do you think is the best neural network currently for processing and possibly generating 44.1 Khz music audio data?
If we're stuck with downsampling to 16 Khz, my question still stands.
- PieSquared 7y agoI don't think anything about the current set of tools is specific to sample rate; WaveNet, Tacotron, WaveRNN, etc, should work fine to generate 44.1Khz audio. They might just need slightly different hyperparameters or sizes to work well, or may take longer to train due to longer sequence lengths.