Équipe Images et Contenus (IC) : MUSHTAQ Umer

Thématiques de recherche : My research focuses on Emotion Recognition in Comics
using multi-modal AI models.

Points forts des activités de recherche : My research focuses on Emotion Recognition in Comics using multi-modal AI models. Comics are a unique narrative medium that convey emotion through a combination of visual art, textual dialogue, facial expressions, and layout design. Recognizing emotions in this context requires interpreting multiple modes of information simultaneously—images, text, and spatial organization. Multi-modal AI are trained to integrate and analyze visual cues (like character expressions or color schemes) alongside textual elements (such as speech balloons or narration) to accurately detect and classify emotions.
Importantly, this opens the door to innovative applications in accessibility for visually disabled individuals, who often miss out on the emotional
nuance of visual storytelling. Multi-modal emotion recognition can help make comics more inclusive, allowing visually impaired readers to experience not just the storyline but also the affective depth of the narrative.