Description
Emotion, or affect recognition is a key to create a truly affect-aware video games and other software. Unfortunately, there are no ready algorithms nor commercial systems available in this area. Thus, intensified research in this field is needed. This project has three main scientific objectives: creation algorithms for affect recognition of video game players, creation of a complete multi-modal affect recognition system to provide affect recognition of video game players and defining a framework that would enable cooperation between developed affect recognition system and any video game. Contribute to development of affect-aware video games is expected as a project outcome. Specific project outputs include multimodal database of emotions, algorithms for affect recognition, framework for cooperation between video games and affect recognition system, a model of player’s emotions and control mechanisms within the game reacting to the detected emotional state of the player and a prototype action-adventure video game cooperating with affect recognition system using defined framework. The target groups are researchers, scientists, and game players.
Summary of project results
Human emotions are an important element of non-verbal human communication and have a great impact on the way of using software, learning or training results, and overall user experience. Therefore, robust and accurate recognition of human emotional states can play an important role in many software systems in different application fields, especially video games. That is why, affective computing has become an emerging research area in recent years. It focuses on development of emotion recognition, interpretation, and processing methods to create affective and affect-aware software to better adapt its behavior to user needs. Emotion recognition is a key to create a truly affect-aware video games and other software. Unfortunately, there are no algorithms or commercial systems available in this area yet. Thus, intensive research in this field is needed. The main goal of Emotion Recognition for Affect-Aware Video Games project (ERAAVG) was to develop emotion recognition algorithms that could be used in affect-aware video games. The project has brought a variety of valuable outcomes including: multimodal Facial Expressions and Emotions Database (FEEDB), KinectRecorder application for data gathering and indexing, a set of algorithms allowing for affect recognition based on depth video stream, which were implemented in prototype EmotiDET application, and Emotion Recognition Framework (ERF) for cooperation between various video games with various emotion recognition algorithms. Additionally, several affect-aware video games have been developed to verify ERF concept and to validate players’ reactions. All the project’s outcomes were disseminated by 14 conference and journal publications and are freely available for other researchers. ERAAVG project was a great chance for scientific development of participating researchers. The achieved results are a solid contribution to affective computing and allowed the development of cooperation with other researchers in that field.
Summary of bilateral results