Text-to-Speech with ESP32 and ZhipuAI

This project uses an ESP32 microcontroller to convert AI-generated text into speech. It utilizes the Baidu TTS API for text-to-speech synthesis and plays the generated audio using an I2S-based audio player. The text-to-speech conversion is performed by sending requests to the ZhipuAI API for generating AI responses.

Features

• Connects to a WiFi network using the ESP32.
• Sends a question to the ZhipuAI API and retrieves a response.
• Converts the AI-generated text into speech using the Baidu TTS API.
• Plays the audio using an I2S audio interface with the MAX98357A DAC.
• Handles audio in small chunks for memory efficiency.

Requirements

• Hardware:

  • ESP32 development board.
  • MAX98357A I2S audio DAC or any compatible I2S-based audio output.
  • Active WiFi connection.

• Software:

  • MicroPython installed on the ESP32.
  • Python packages: urequests, ujson, gc, network, and machine.

• APIs:

  • ZhipuAI (for AI text generation).
  • Baidu TTS (for text-to-speech conversion).

Setup

1. Wi-Fi Configuration:

   • Replace 'xxx' in the SSID and PASSWORD constants with your actual WiFi credentials.

2. API Keys:

   • Replace 'xxx' in the API_KEY variable with your ZhipuAI API key.
   • Replace 'xxx' in the api_key and secret_key fields inside the main() function with your Baidu TTS API credentials.

3. I2S Audio Configuration:

   • The code is configured to use an I2S connection to a MAX98357A DAC. You can modify the SCK_PIN, WS_PIN, and SD_PIN variables to suit your hardware.
   • The audio sample rate is set to 8000Hz (you can change it to 16000Hz or 24000Hz depending on your needs).

How to Use

1. Flash the MicroPython firmware onto the ESP32 board if you haven't already.
2. Upload this Python script to the ESP32 using any appropriate method, such as Thonny or WebREPL.
3. Connect the ESP32 to a serial terminal or a REPL interface.
4. Run the script. It will first connect to the WiFi and prompt you to enter a question for the AI.
5. The ESP32 will:
   • Send the question to the ZhipuAI API and receive an AI-generated response.
   • Convert the response text to speech using the Baidu TTS API.
   • Play the generated audio in chunks through the connected I2S DAC.

Example Interaction