Build a Voice‑Enabled Chatbot in Python with Baidu AI and Qingyunke

Brief Overview

In the past two days I needed to develop a small Python program that enables a human‑machine conversation using an intelligent dialogue interface. The goal is to achieve voice‑based interaction where a user can speak, the system recognizes the speech, obtains a textual query, sends it to a chatbot, and finally plays the chatbot's answer as speech.

Overall Idea

Computer receives the user's voice input .

The voice input is converted into text .

The text is sent to an intelligent dialogue API, which returns a smart answer in text .

The answer text is converted back to voice format for playback.

Required Environment

pip install pyaudio for recording and generating wav files

pip install baidu-aip Baidu AI SDK for speech‑to‑text

pip install pyttsx3 for converting text to speech

Receive User Voice Input and Save as Audio File

import time</code><code>import wave</code><code>from pyaudio import PyAudio, paInt16</code><code>framerate = 16000  # sample rate</code><code>num_samples = 2000  # frames per buffer</code><code>channels = 1</code><code>sampwidth = 2  # 2 bytes</code><code>FILEPATH = '../voices/myvoices.wav'  # ensure directory exists</code><code>class Speak():</code><code>    # Save audio data to wav file</code><code>    def save_wave_file(self, filepath, data):</code><code>        wf = wave.open(filepath, 'wb')</code><code>        wf.setnchannels(channels)</code><code>        wf.setsampwidth(sampwidth)</code><code>        wf.setframerate(framerate)</code><code>        wf.writeframes(b''.join(data))</code><code>        wf.close()</code><code>    # Record voice</code><code>    def my_record(self):</code><code>        pa = PyAudio()</code><code>        stream = pa.open(format=paInt16, channels=channels, rate=framerate, input=True, frames_per_buffer=num_samples)</code><code>        my_buf = []</code><code>        t = time.time()</code><code>        print('正在讲话...')</code><code>        while time.time() < t + 5:  # record for 5 seconds</code><code>            string_audio_data = stream.read(num_samples)</code><code>            my_buf.append(string_audio_data)</code><code>        print('讲话结束')</code><code>        self.save_wave_file(FILEPATH, my_buf)</code><code>        stream.close()

Call Baidu AI Interface for Speech Recognition

First, create an account on the Baidu AI Open Platform, enable the Speech Recognition service, and obtain AppID , API Key , and Secret Key . These credentials are used to call the API.

After the credentials are ready, the following code sends the recorded wav file to Baidu for recognition.

from aip import AipSpeech</code><code>APP_ID = '25990397'</code><code>API_KEY = 'iS91n0uEOujkMIlsOTLxiVOc'</code><code>SECRET_KEY = ''  # fill your secret key</code><code>client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)</code><code>class ReadWav():</code><code>    def get_file_content(self, filePath):</code><code>        with open(filePath, 'rb') as fp:</code><code>            return fp.read()</code><code>    def predict(self):</code><code>        return client.asr(self.get_file_content('../voices/myvoices.wav'), 'wav', 16000, {'dev_pid': 1537})</code><code>readWav = ReadWav()</code><code>print(readWav.predict())

Sample output:

{'corpus_no': '7087884083428433929', 'err_msg': 'success.', 'err_no': 0, 'result': ['你叫什么名字呀？'], 'sn': '255158586831650276613'}

Request Intelligent Robot and Get Reply

A free, no‑registration chatbot service called Qingyunke can be called with a simple GET request.

def talkWithRobot(msg):</code><code>    url = 'http://api.qingyunke.com/api.php?key=free&appid=0&msg={}'.format(urllib.parse.quote(msg))</code><code>    html = requests.get(url)</code><code>    return html.json()["content"]</code><code>print(talkWithRobot("你好呀!"))

Typical response: 哟~ 都好都好

Convert Answer to Speech and Play

import pyttsx3</code><code>class RobotSay():</code><code>    def __init__(self):</code><code>        self.engine = pyttsx3.init()</code><code>        self.rate = self.engine.getProperty('rate')</code><code>        self.engine.setProperty('rate', self.rate - 50)</code><code>    def say(self, msg):</code><code>        self.engine.say(msg)</code><code>        self.engine.runAndWait()</code><code>robotSay = RobotSay()</code><code>robotSay.say("你好呀")  # plays the text as speech

Combine Everything into an Automatic Chatbot

def talkWithRobot(msg):</code><code>    url = 'http://api.qingyunke.com/api.php?key=free&appid=0&msg={}'.format(urllib.parse.quote(msg))</code><code>    html = requests.get(url)</code><code>    return html.json()["content"]</code><code>robotSay = RobotSay()</code><code>speak = Speak()</code><code>readTalk = ReadWav()</code><code>while True:</code><code>    speak.my_record()  # record voice</code><code>    text = readTalk.predict()['result'][0]  # speech‑to‑text</code><code>    print("本人说:", text)</code><code>    response_dialogue = talkWithRobot(text)  # chatbot reply</code><code>    print("青云客说:", response_dialogue)</code><code>    robotSay.say(response_dialogue)  # play reply

Sample interaction:

正在讲话...</code><code>讲话结束...</code><code>本人说: 你好呀。</code><code>青云客说: 哟~ 都好都好</code><code>...

Future Work

The current implementation is a simple command‑line prototype. Next steps include building a GUI, adding more features, and sharing the complete project with the community.