4 ms·
Machine learning for signal processing is no joke. It's an order of magnitude improvement. The approach is different, too. Instead of approaching the signal mat
by echelon 6y ago
Machine learning for signal processing is no joke. It's an order of magnitude improvement. The approach is different, too. Instead of approaching the signal mathematically with Fourier transforms and such, you approach it as part of a pipeline where you extract high level components such as pitch and timber. The ML model learns how to do that.
This field is going to change everything. Look what I've been able to do with some of the recent ML innovations: https://vo.codes https://vo.codes
- nullsense 6y agoThe next layer that would put the realism off the charts is ML that's able to replicate a given speakers idiolect, not just their voice. Mimicking people's ideolects is a hobby of mine, and I typed in phrases that Dr Phil would say, and the voice itself is good but the delivery is off enough that I know it's definitely not him. Im sure it all get there one day.