src/speech.py

from config import settings
from playsound import playsound
from typing import List
from src.metahuman import MetaHuman
from src.stt_api import transcribe_file


helper_path = settings.data_path / "MetaHuman_Voice"
user_path = settings.data_path / "User1_Voice"


def run_text(helper):

    response = None

    while True:

        audio_file = helper.flow(response=response)

        playsound(audio_file, block=False)
        print(f"Helper: {helper.get_state_transcript()}")

        response = input("Enter response: ")
        print(f"User: {response}")
        if response.lower() == "stop":
            break


def run_robouser(helper, user_response_flow: List[str]):

    helper.flow()
    print(f"Helper: {helper.get_state_transcript()}")

    while user_response_flow:

        user_audio_file = user_path / user_response_flow.pop(0)
        print("User:")
        user_response = transcribe_file(user_audio_file)

        helper.flow(user_response)
        print(f"Helper: {helper.get_state_transcript()}")


if __name__ == "__main__":

    from model import ModelCompare
    
    helper = MetaHuman(helper_path, compare_func=ModelCompare())
    run_text(helper)

    user_response_flow = ["Block_okidoki.wav", "Block_yep.wav",
                          "Block_service.wav", "Block_production.wav", 
                          "Block_back.wav", "Block_menu.wav"]
    run_robouser(helper, user_response_flow)