Emotion Recognition Using a Robocentric Perspective in Indoor Environments

• Leonardo B. Gamarra Portal Universidad Catolica San Pablo
• Raquel E. Patino-Escarcina Universidad Catolica San Pablo

Resumo

Nowdays robots have been inserted in various areas of society. For this reason they need answer to recognize their environment and interact with other agents as people or other robots. In this sense, emotion recognition is very important to decide right actions of the robots in their environment. This research addresses the importance of emotion recognition to achieve human-robot interaction, which is a fundamental and decisive element in the area of social robotics. Emotion recognition is implemented using the DeepFace library, which incorporates five face detectors and seven convolutional neural networks for facial recognition and emotion analysis.

Palavras-chave: Emotion Recognition, Human-Robot Interaction(HRI), Convolutional Neural Network(CNN), Robot Operating System(ROS)