音频文字转写

import speech_recognition as sr
from pydub import AudioSegment
import urllib.request
import os
# 设置音频文件的位置
# Load the audio file
current_dir=current_dir = os.path.dirname(os.path.abspath(__file__));
file_path = os.path.join(current_dir, '', 'girl.mp3') 
# 下载远程文件到本地
#url = "https://yewen0112.oss-cn-shanghai.aliyuncs.com/video-mini/audio_20230417.mp3"
#res =urllib.request.urlretrieve(url, file_path)

# Convert to WAV format
audio_file = AudioSegment.from_mp3(file_path)
file_path_wav = os.path.join(current_dir, '', 'audio.wav')  
audio_file = audio_file.set_frame_rate(16000)
audio_file = audio_file.set_channels(1)
audio_file = audio_file.set_sample_width(2)
audio_file_wav = audio_file.export(file_path_wav, format="wav")

# 创建 SpeechRecognition 对象
r = sr.Recognizer()

# 读取音频文件
with sr.AudioFile(file_path_wav) as source:
    audio = r.record(source)

# 识别音频文件
try:
    record=r.recognize_google(audio, language='zh-CN')
    print(record)
    learning_quality_keywords = ['恭喜', '发财']
    # 匹配关键词
    learning_quality_count = sum(keyword in record for keyword in learning_quality_keywords)
    print(learning_quality_count)
except sr.UnknownValueError:
    raise 'Google Speech Recognition could not understand audio'
except sr.RequestError as e:
    print("Could not request results from Google Speech Recognition service" + format(e))

import os
import whisper
import opencc 
# 设置音频文件的位置
current_dir=current_dir = os.path.dirname(os.path.abspath(__file__));
file_path = os.path.join(current_dir, '', 'girl.mp3')  

#选择模型medium
model = whisper.load_model("medium")

result = model.transcribe(file_path)
record=result["text"]
cc = opencc.OpenCC('t2s')
record=cc.convert(record)
print(record)
learning_quality_keywords = ['恭喜', '发财']
# 匹配关键词
learning_quality_count = sum(keyword in record for keyword in learning_quality_keywords)
print(learning_quality_count)