dragonfly_speech_recognition

ROS Wrapper for https://github.com/dictation-toolbox/dragonfly.

Prerequisites server

• Windows speech recognition engine (Windows Vista+)
• Python (https://matthewhorne.me/how-to-install-python-and-pip-on-windows-10/)
• PyWin32 (pip install pywin32 or use the installer) (https://github.com/mhammond/pywin32)
• PyYaml (pip install pyyaml)
• Git (https://gitforwindows.org/)

Installation

git clone https://github.com/tue-robotics/dragonfly_speech_recognition.git --recursive
cd dragonfly_speech_recognition/src/dragonfly_speech_recognition/dragonfly
pip install -e .

Post Install

• Disable firewall
• Set static IP
• Disable sleep & screen saver

Autostart:

• goto to startup folder by entering the following in run: shell:startup
• create start_dragonfly_speech_recognition.bat:

  :loop
  python -u "C:\Users\amigo\dragonfly_speech_recognition\scripts\dragonfly_server" --ip 192.168.44.XX
  goto loop
  

Troubleshooting

version = p.load()
EOFError

Remove the following folder on your windows machine: C:\Python27\Lib\site-packages\win32com\gen_py

Client installation

Assuming a ROS environment has already been setup, clone the package in your workspace

git clone https://github.com/tue-robotics/dragonfly_speech_recognition.git
catkin_make/catkin build