python_語音轉文字_參考程式

from google.colab import drive
drive.mount('/content/drive')

!pip install moviepy
!pip install git+https://github.com/openai/whisper.git
!pip install opencc-python-reimplemented

import whisper
from opencc import OpenCC  # 用於簡體轉繁體中文
import os

# Step 1: 使用 Whisper 將音頻轉換為文字
def transcribe_audio(audio_file):
    model = whisper.load_model("medium")  # 使用 Whisper 模型
    result = model.transcribe(audio_file, language='zh')  # 語言設置為中文
    return result['text']

# Step 2: 將簡體中文轉為繁體中文
def simplify_to_traditional(text):
    cc = OpenCC('s2t')  # s2t 表示簡體轉繁體
    return cc.convert(text)

# Step 3: 完整流程：m4a -> 繁體中文文字
def m4a_to_text(m4a_file, output_folder):
    # 語音轉文字
    transcript = transcribe_audio(m4a_file)
    
    # 簡體轉繁體
    traditional_text = simplify_to_traditional(transcript)
    
    # 將文字儲存為文本文件
    text_output_file = os.path.join(output_folder, 'transcription.txt')
    with open(text_output_file, 'w', encoding='utf-8') as f:
        f.write(traditional_text)
    
    return traditional_text

# 您的 .m4a 文件路徑和輸出資料夾
data_path = '/content/drive/My Drive/12345678/12345678_file.m4a'
output_folder = '/content/drive/My Drive/12345678/'

# 開始處理
transcribed_text = m4a_to_text(data_path, output_folder)
print(transcribed_text)