Multimodal Emotion Recognition

My PhD thesis on audio & video based emotion recognition

My Ph.D. dissertation addressed the research problem of automatic multimodal emotion recognition using audio and visual cues. It approached this problem through several methods and proposed techniques to exploit audio and video channels’ complementary and supplementary information. The studies conducted in my Ph.D. work targeted several aspects of multimodal emotion recognition, such as investigating various representations of audio and visual cues, proposing fusion algorithms, utilizing the temporal information in a video clip to integrate both modalities and studying the patterns and behavior of audio-video signals for emotion recognition as a function of time. You can access my Ph.D. dissertation from the following link:

The following figure shows the cover of my PhD dissertation. The cover was designed by Gizem Ustuner.