Emotion Assessment and Adaptation of User Interfaces: A Deep Learning Approach

Résumé

<div><p>Human-Computer Interaction (HCI) plays a crucial role in designing interfaces that enhance user experiences by making technology more intuitive, efficient, and responsive to human needs. This dissertation presents three significant contributions to the fields of HCI and emotion recognition, aiming to further this goal. Our first proposed approach involves the development of a comprehensive computational model that integrates facial expressions, speech, and visual-audio data. This model was trained on an open-source dataset and subsequently fine-tuned to enhance its performance. The fine-tuned model was then deployed within the HCI domain to assess affective responses to various graphical user interfaces (GUIs), distinguishing between good and bad designs. This evaluation utilized multimodal data, including facial expressions, eye tracking (ET), and electroencephalography (EEG) signals, to provide a holistic understanding of user reactions. The second proposed approach is the fine-tuning of this computational model using data collected from a real-world experiment involving diverse users. This experiment aimed to capture authentic affective responses and refine the model for better accuracy and reliability in predicting emotional states. The last contribution is the integration of the refined computational model into an emotional adaptive user interface (UI) system. This system dynamically adjusts the UI based on real-time emotional feedback from users, creating a more personalized and responsive interaction experience. Notably, this study is the first to demonstrate the effectiveness of a personalized computational model in adapting UIs within the HCI domain.</p><p>The results indicate that computational models can significantly enhance user experience by providing adaptive interfaces that respond to individual emotional states. This research highlights the potential of computational models to revolutionize the field of HCI by enabling the creation of emotion-sensitive UIs paving the way for more intuitive and user-centered interface adaptations.</p></div>

Citer ce document

Accès au document

Ce document est hébergé sur son dépôt institutionnel d'origine.

Auteur(s)

Statistiques

Consultations : 2

Téléchargements : 0