Experimental emotion recognition system and services for mobile network environments

The two state (pleasant and unpleasant) emotion recognition engine was developed for mobile phone, and experimental demo system was developed. To recognize the emotion of the mobile phone user, sensor enabled watch was developed, and it transfers bio signals and environmental information to the emotion recognition engine located at the mobile phone through wireless personal area network. In experimental system, the emotion recognition engine decides the emotion of the user, and transfers the result to the service enablers at Internet through cellular network. Emotive service enablers provide differentiated multimedia contents based on the emotion of the user through IPTV, mobile social network service, and blog service. To increase the accuracy of the emotion recognition, compensation methods were applied for the sensor enabled watch. According to our emotion decision algorithm, the accuracy was evaluated as 82% for experimental persons who participated in the test.