HN Explorer

Ask HN: On-Device Text to Speech

3 points by jamil7 ↗ HN
Hey HN, has anyone found a viable solution for doing this locally and offline on iOS? I'd like to offer a privacy-friendly text to speech feature to my App, and Apple's speech synthesis sounds awful compared to some newer models and TTS engines. The only thing I've found is an older TensorFlowTTS example here: https://github.com/TensorSpeech/TensorFlowTTS/tree/master/examples/ios

Any pointers or tips appreciated.

4 comments

[ 4.7 ms ] story [ 23.7 ms ] thread
layer8
Did you take a look at the iOS 17 beta? With the new Personal Voice feature, I would expect an improvement in speech synthesis. Furthermore, users will maybe expect to be able to use their personal voices with your app.
jamil7
Thanks for the reply, I did, but it didn't seem like something I could directly integrate with from within my app, I'll dig into it a bit more.
layer8
I haven’t looked much into it myself, but https://developer.apple.com/videos/play/wwdc2023/10033/ seems to describe how you can (a) write a speech synthesis provider (if indeed you want to integrate a custom voice) and (b) use such a “personal voice” in an app. Regarding the latter part, I would assume that the AI-provided voices will be relatively high-quality.
jamil7
Thanks for that, I ended up messing around with this on the iOS17 Beta and, in theory it seems like a lot has improved but the betas seem to be missing a lot of the actual voices.