Social Mind

Overview

Social HRI Software Framework - Social Mind - is the modularized HRI software framework for developing the social robots. The framework is modularized by each function such as sensory perception, social perception, social mind, dialog, motion arbiter and motion renderer. Each module sends and receives the information (event, data) with each other using the ROS topic, service and actionlib.

Installation (by sources)

    $ cd ~/catkin_ws/src
    $ git clone https://github.com/byeongkyu/social_mind.git
    $ rosdep install --from-paths social_mind --ignore-src -r -y
    $ cd social_mind
    $ find -name 'requirements.txt' | xargs -L 1 sudo pip install -U -r 

Topics

/raising_events RaisingEvents.msg via events_multiplexer

• Type

    Header header
    string recognized_word
    string[] events
    string[] data
  

  • Example: Speech Recognition

    header:
        seq: 12
        stamp:
        secs: 1505344152
        nsecs: 513442993
        frame_id: ''
    recognized_word: Hello
    events: ['speech_recognized']
    data: ['{"confidence": 1.0, "recognized_word": "Hello"}']
    

  • Example: Button Input

    header:
        seq: 52
        stamp:
        secs: 1505344319
        nsecs:  13427972
        frame_id: ''
    recognized_word: ''
    events: ['button_pressed']
    data: ['{"text": "yes"}']
    

/reply Reply.msg received from Scheduler Engine

• Type

    Header header
    string reply
  

• Format

  • Say

    header:
        seq:
        stamp:
        secs:
        nsecs:
        frame_id: ''
    reply: "Hello. My name is Silbot"
    

  • Say with gesture by tag

    header:
        seq:
        stamp:
        secs:
        nsecs:
        frame_id: ''
    reply: "<sm=tag:neutral> Hello. My name is Silbot"
    

  • Say with gesture by specific motion

    header:
        seq:
        stamp:
        secs:
        nsecs:
        frame_id: ''
    reply: "<sm=play:sm/open/1> Hello. My name is Silbot"
    

  • Say with expession

    header:
        seq:
        stamp:
        secs:
        nsecs:
        frame_id: ''
    reply: "<expression=happiness> Hello. My name is Silbot"
    

Tag and Gesture List for Robots typically can find in each robot's render_gesutre package: motions.yaml