My Current Research Work

Affective computing emerges as a critical component in enhancing AI agents with emotional intelligence, centered on understanding how human emotions impact decision-making, behavior, and communication. The study focuses on enhancing this field by investigating different models, fusion techniques, and data preprocessing methods to determine the most effective ones for emotion recognition in a transparent manner. The primary aim is to create a model for recognizing emotions using audio and video data, relying on deep neural networks like Transformers. Furthermore, delving into the potential of applying transfer learning using pre-existing models for this specific task is a key aspect of the research. The model's performance is evaluated on the IEMOCAP dataset, emphasizing accuracy and F-score. The goal is to shed light on the best approaches for replicating and comprehending the emotional facets of human intelligence in intelligent systems.