python调用百度语音识别api python调用百度语音识别api（基于百度ai的语音识别python）

import requests
import json
import uuid
import base64

def get_token():
 url = "https://openapi.baidu.com/oauth/2.0/token"
 grant_type = "client_credentials"
 api_key = "NzGBYD0jPFDqVT8VHRYa****"  # 自己申请的应用
 secret_key = "8439155b9db2040b4acd13b0c*****" # 自己申请的应用
 data = {'grant_type': 'client_credentials', 'client_id': api_key, 'client_secret': secret_key}
 r = requests.post(url, data=data)
 token = json.loads(r.text).get("access_token")
 return token


def recognize(sig, rate, token):
 url = "http://vop.baidu.com/server_api"
 speech_length = len(sig)
 speech = base64.b64encode(sig).decode("utf-8")
 mac_address = uuid.UUID(int=uuid.getnode()).hex[-12:]
 rate = rate
 data = {
 "format": "wav",
 "lan": "zh",
 "token": token,
 "len": speech_length,
 "rate": rate,
 "speech": speech,
 "cuid": mac_address,
 "channel": 1,
 }
 data_length = len(json.dumps(data).encode("utf-8"))
 headers = {"Content-Type": "application/json",
 "Content-Length": data_length}
 r = requests.post(url, data=json.dumps(data), headers=headers)
 print(r.text)


filename = "two.wav"

signal = open(filename, "rb").read()
rate = 8000

token = get_token()
recognize(signal, rate, token)

import scipy.io.wavfile
filename = "two.wav"
rate, signal = scipy.io.wavfile.read(filename=filename)