sherpa-onnx: Speech to text and text to speech running locally on CPU

[csukuangfj]

[sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx). `[Delphi]` `[FPC]`

It provides Object Pascal APIs for speech to text and text to speech. Everything runs locally on CPU.
Support commonly used models, such as Zipformer, Whisper, Paraformer, and SenseVoice for speech
recognition, silero VAD for voice activity detection, VITS for text to speech.

Examples for object pascal API can be found at https://github.com/k2-fsa/sherpa-onnx/tree/master/pascal-api-examples
Examples for Lazarus can be found at https://github.com/k2-fsa/sherpa-onnx/tree/master/lazarus-examples
Doc is available at

• https://k2-fsa.github.io/sherpa/onnx/lazarus/index.html
• https://k2-fsa.github.io/sherpa/onnx/pascal-api/index.html

Note that it is actively maintained and it uses Apache 2.0 license.