This package enables you to demonstrate "denden-mushi", which picks up the sound with the microphone, analyzes the emotion using HUME API and expresses it with its eyes and cheek color.
- M5Stack用NeoPixel互換LED搭載 HEXボード: https://www.switch-science.com/products/6058?_pos=1&_sid=04326fe49&_ss=r
- radxa zero
- ReSpeaker 4 Mic Array (UAC1.0)
- M5 ATOM ECHO
- M5Stack Fire
- Round Display Module with M5Stamp C3 (pio env name: stampc3) : https://www.switch-science.com/products/8098
To start the program, do:
roslaunch jishupuro_pkg demo.launch
You can test if devices are activated by doing this:
rostopic pub -1 /eye_status std_msgs/UInt16 "data: 6"
Now, visit http://<radxa's ip>:8085/jishupuro_pkg/hume_control_getip.html and fill in the blank with your radxa's IP adress, press "Set IP Address" button. If it doesn't work on Chrome, try on Safari.
You can start the program by pressing "Start Hume", and stop it by "Stop Hume".
Get your API key at https://dev.hume.ai/intro
nano ~/.bashrc
Append export HUME_API="your_api_key"
source ~/.bashrc
You may have to do these to execute the program.
sudo apt-get install ros-noetic-roswww
cd ~/catkin_ws/src
git clone --single-branch -b stereo-semi https://github.com/iory/jsk_demos
These websites will help you to rename ports of the devices.
- https://smdn.jp/electronics/udev_create_persistent_usb_device_symlink/ ,
- https://docs.google.com/presentation/d/1990HMAx-kUocaGTCPdmen9bIHAOFU23R76boXbFtALY/edit#slide=id.g25e40fad838_2_50 (p13)
wget https://gist.githubusercontent.com/iory/ecd4b6c19fd657df0979a96dab6c3d56/raw/936b0b3d1db1d669efb5e711b725ea7b363730a3/pairing_bt.py
python3 pairing_bt.py
To see the initial setup for python3.8 virtual environment, visit https://www.linuxcapable.com/how-to-install-python-3-8-on-debian-linux/#google_vignette .
write
source ~/catkin_ws/devel/setup.bash in bashrc so that you do not need to do it every time.