iCub: Learning Emotion Expressions using Human Reward

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Workshop on Bio-inspired Social Robot Learning in Home Scenarios - Oct 2016.
Associated documents : Churamani_IROS2016_BSRL-HS.pdf [3.2Mo]  
The purpose of the present study is to learn emotion expression representations for artificial agents using reward shaping mechanisms. The approach takes inspiration from the TAMER framework for training a Multilayer Perceptron (MLP) to learn to express different emotions on the iCub robot in a human-robot interaction scenario. The robot uses a combination of a Convolutional Neural Network (CNN) and a Self-organising Map (SOM) to recognise an emotion and then learns to express the same using the MLP. The objective is to teach a robot to respond adequately to the user's perception of emotions and learn how to express different emotions.

 

@InProceedings\{CCGB16,
  author       = "Churamani, Nikhil and Cruz, Francisco and Griffiths, Sascha and Barros, Pablo",
  title        = "iCub: Learning Emotion Expressions using Human Reward",
  booktitle    = "IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Workshop on Bio-inspired Social Robot Learning in Home Scenarios",
  month        = "Oct",
  year         = "2016",
  address      = "Daejeon, KR",
  url          = "https://www2.informatik.uni-hamburg.de/wtm/publications/2016/CCGB16/"
}

» Nikhil Churamani
» Francisco Cruz
» Sascha Griffiths
» Pablo Barros