Skip to main content

5.2.8 Text to Speech

Introduction​

This section describes how to convert a text into a speech and play it through an audio output interface.

Code repository: (https://github.com/D-Robotics/hobot_tts.git)

Supported Platforms​

PlatformSystemFunction
RDK X3, RDK X5Ubuntu 20.04 (Foxy), Ubuntu 22.04 (Humble)Subscribe to text messages, convert them into speech, and play them out

Note: Only supports RDK X3, RDK X3 Module is not supported yet.

Preparation​

RDK​

  1. RDK has been flashed with Ubuntu 20.04/22.04 system image provided by D-Robotics.
  2. TogetheROS.Bot has been successfully installed on RDK.
  3. An audio driver board compatible with D-Robotics has been obtained, and the environment has been set up according to Smart Voice section
  4. Connect the audio board's headphone interface with headphones or speakers.

Usage​

RDK​

  1. For the first run, you need to download the models file and extract it. The commands are as follows:

    # Configure the tros.b environment
    source /opt/tros/setup.bash
    wget http://archive.d-robotics.cc/tros/tts-model/tts_model.tar.gz
    sudo tar -xf tts_model.tar.gz -C /opt/tros/${TROS_DISTRO}/lib/hobot_tts/

  2. Run the following command to check if the audio device is normal:

    root@ubuntu:~# ls /dev/snd/
    by-path  controlC0  pcmC0D0c  pcmC0D1p  timer

    If a similar audio playback device like pcmC0D1p appears, it means the device is working fine.

    caution

    When using the audio board for the first time, you need to use srpi-config for configuration, otherwise the audio device will not be recognized.

    For the configuration method, refer to the RDK User Manual Audio Adapter Board

  3. Start the hobot_tts program.

    # Configure the tros.b environment
    source /opt/tros/setup.bash
    # Disable debug information
    export GLOG_minloglevel=1

    ros2 run hobot_tts hobot_tts

Note: If the audio playback device is not pcmC0D1p, you need to use the playback_device parameter to specify the playback audio device. For example, if the audio playback device is pcmC1D1p, the launch command should be: ros2 run hobot_tts hobot_tts --ros-args -p playback_device:="hw:1,1"

  1. Open a new terminal and use the echo command to publish a topic
# Configure the tros.b environment
source /opt/tros/setup.bash
ros2 topic pub --once /tts_text std_msgs/msg/String "{data: ""Do you know the horizon? Yes, I know the horizon. It is a line that extends from the ground to the sky, defining the boundary between the ground and the sky.""}"
  1. You can hear the playback sound from your headphones or speakers.

Notes​

Currently, only Chinese and English text content are supported. Do not publish text messages in other languages.

Last updated on