The researchers undertaking this project aim to develop several emotionally intelligent ECAs that can recognize and respond effectively to user affect. The project begins with a selection of application areas for which it may be possible to build an ECA. These application areas may include but are not limited to educational software, games, productivity tools, business tools and others. Having selected the target domain or domains and the target audience, the ECA’s physical appearance will then be designed. Several designs will be tested using a Wizard-of-Oz technique to determine which design appeals most to the users. Concurrently, the researchers will conduct focus group discussions and observations of human-to-human and human-to-machine interactions to determine appropriate responses to certain user affective states. After a design has been selected, the models of user affect will be integrated with the ECA. This emotionally intelligent ECA will then be tested again to determine whether it indeed recognizes user affective states, whether it responds appropriately and how it’s responses affect users.
Funded by Philippine Council for Industry, Energy and Emerging Technology Research and Development (PCIEERD)
Related Projects
Controlling a Robotic Hand Using a Non-Invasive Brain-Computer Interface
Continue reading →
Developing an Affect Sensitive Horror Game
Continue reading →
Developing an Android-Based Game Tool with Monitoring Support to Track Stroke Patient Progress During Rehabilitation
Continue reading →
Exploring a System for Automatically Detecting Driver Anxiety in a Virtual Racing Game
We describe a work-in-progress that attempts to detect anxiety in drivers while exposed to a virtual racing game. The detection system is a combination of two Nintendo Wii Balance Boards, parallel video recording of in-game events and facial expressions, and the OCZ Technology’s Neural Impulse Actuator (NIA) brain-controlled headset.
Continue reading →
