[Core] Add better support for Text-To-Speech - Githubissues

Harium / suneidesis

A project to represent knowledge

MIT License

2 stars 1 forks source link

[Core] Add better support for Text-To-Speech #137

Open yuripourre opened 2 years ago

yuripourre commented 2 years ago

Currently, is possible to have an output that converts text to a sound but there is no way to set voice parameters in a standard way.

One way to solve/improve this is to add voice parameters (hints) to the context map and parse the text looking for SSML tags.

Related to #109

yuripourre commented 2 years ago

Let's build a proof of concept?

Create emulator
Install Google Speech Services download the apk from external sources
Run emulator without UI (https://stackoverflow.com/a/63694326)
Run app from command line passing param (the sentence you want to speak)(https://stackoverflow.com/a/24322789)

TTS from Android emulator (should read a custom protocol) (https://stackoverflow.com/a/68869600)

Compile and install APK in running emulator (https://developer.android.com/studio/build/building-cmdline.html)

Custom protocol, maybe tts://sentence?text=something%20is%20wrong https://stackoverflow.com/a/24032099 https://github.com/hkusoft/AndroidHowto/blob/master/CustomUrl/app/src/main/java/studio/xy/customurl/MainActivity.java