Open mablue opened 4 months ago
Sorry. TTS Server is only for calling network or local TTS, I have no plans for a built-in offline engine.
Sherpa's biggest drawback is its limited GUI, which doesn't allow users to install multiple language dictionaries and rules. However, it excels as an offline Microsoft Read Aloud alternative.
TTS-server's main concern is its potential vulnerability to becoming outdated if technology advancements are not closely monitored, rendering its online services obsolete. It's important to remember that Microsoft Read Aloud is a new service and shouldn't be dismissed prematurely.
Combining these two technologies has the potential to create the best accessibility tool ever designed for the visually impaired. I plan to discuss this possibility with @csukuangfj, who might be interested in integrating this technology into their GUI.
While I believe you are an excellent Android developer, I consider @csukuangfj to be the superior expert in { #onnx, #pytorch } AI programming for this specific task.
If the sherpa-onnx.apk supports multiple models, would it be possible to use the local TTS functionality of the TTS server to call it?
If the sherpa-onnx.apk supports multiple models, would it be possible to use the local TTS functionality of the TTS server to call it?
Sure, The Sherpa team is seeking an expert contributor for Android UI development for make a multi voice/language UI for sherpa. Additionally, the tts-server may benefit from the incorporation of AI functionality.
Hi. Thanks for your support and fixs. We as persian blinds community now have not any bug in latest test version :)
Please Add sherpa onnx tts models support.
Its an small opensource apk that can read sherpa models If we can support these models internally inside tts-server we can mix up different offline languages internally in tts server and use them for multilingual porpuses. Them are like online microsoft service but them dont need internet connection that can be useable for people that some times have not internet connection. This is the link of compiled apk files https://k2-fsa.github.io/sherpa/onnx/tts/apk-engine.html
Also we have hugging face directory for compile and generate the tts engines apk automatically:
https://huggingface.co/csukuangfj/sherpa-onnx-apk/tree/main/tts-engine