M5Stack ModuleLLMでGeminiAPIを使うスクリプト

gemini_arecord.py

import subprocess
import speech_recognition as sr
from gtts import gTTS
import os
import google.generativeai as genai

# ==== 設定 ====
GEMINI_API_KEY = "YOUR_API_KEY"  # ← あなたのAPIキーに変更
genai.configure(api_key=GEMINI_API_KEY)
model = genai.GenerativeModel(model_name="models/gemini-2.0-flash")

# ==== 録音（arecord） ====
def record_with_arecord(filename="input.wav", duration=10, device="plughw:0,0"):
    print("🎙 arecordで録音中...")
    subprocess.run(["arecord", "-D", device, "-f", "cd", "-t", "wav", "-d", str(duration), "-r", "16000", filename])

# ==== 音声認識 ====
def recognize_from_file(filename="input.wav"):
    r = sr.Recognizer()
    with sr.AudioFile(filename) as source:
        audio = r.record(source)
    try:
        return r.recognize_google(audio, language="ja-JP")
    except sr.UnknownValueError:
        return None
    except sr.RequestError as e:
        print(f"🔌 Google音声認識エラー: {e}")
        return None

# ==== Geminiへの問い合わせ ====
def ask_gemini(user_prompt, system_prompt):
    full_prompt = system_prompt + "\n" + user_prompt
    response = model.generate_content(full_prompt)
    return response.text

# ==== 音声合成と再生 ====
def speak_text(text):
    print("🗣 gTTSで音声合成...")
    tts = gTTS(text=text, lang="ja")
    tts.save("response.mp3")
    subprocess.run(["ffmpeg", "-y", "-i", "response.mp3", "-ar", "44100", "response.wav"],
                   stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
    subprocess.run(["aplay", "response.wav"])
    os.remove("response.mp3")
    os.remove("response.wav")

# ==== メインループ ====
def main():
    device = "plughw:0,0"
    system_prompt = "あなたは可愛い手乗りロボットです。回答は語尾に「ござる」を付けて３０秒以内に収めてください。"
    print("✅ Ctrl+C で終了します\n")

    while True:
        record_with_arecord(device=device)
        user_text = recognize_from_file()
        if user_text:
            print(f"📝 認識: {user_text}")
            reply = ask_gemini(user_text, system_prompt)
            print(f"🤖 Gemini: {reply}")
            speak_text(reply)
        else:
            print("❌ 音声が認識できませんでした")

if __name__ == "__main__":
    main()

インストール.txt

# 他にもあるかもしれませんが、エラーを生成AIに聞くなどしてください。（教えていただけると助かりますm(_ _)m)

apt install ffmpeg portaudio19-dev alsa-utils flac python3-pip 



pip install pyaudio speechrecognition gtts playsound google-generativeai