Author: Antonio Escamilla Pinilla
Programme: Doctoral Programme in Network and Information Technologies
Idioma: English
Supervision: Dr Javier Melenchón Maldonado and Dr Carles Monzo Sánchez
Faculty / Institute: Doctoral School UOC
Subjects: Computer Science
Key words: human-centered computing, interaction design, empirical study in interaction design, motion-based feature, feature visualization
Area of knowledge: Network and Information Technologies
Summary
Human-computer interaction (HCI) is evolving towards more natural, social, and user-centered interactions, specifically focusing on the use of the entire body as an interface. Full-body interaction involves capturing and analyzing body movements to control and manipulate digital content, enabling immersive and engaging experiences. In this context, computer vision technology is preferred over add-ons, wearables, or marker systems due to its non-intrusive, cost-effective, and versatile nature. By eliminating the need for additional devices, it allows for a more natural and unrestricted observation of human motion, ensuring a genuine representation of movement. However, designing effective full-body interaction systems requires understanding of human motion and its interpretation, which poses challenges for interaction designers. This thesis addresses the critical need for structured relationships between human motion and feature extraction technology, which arises from the problem that non-technically skilled practitioners often struggle to utilize motion-based features effectively. To provide better context, the thesis presents the challenges faced by these professionals in working with complex technical parameters and motion-based features, which hinder their ability to create multi-user interactive experiences. By highlighting this issue early on, the thesis aims to emphasize its significance and underscore that its primary objective is not just to resolve interaction design challenges but also to offer the additional value of empowering non-technical profiles by facilitating their design work through accessible motion feature extraction technology. In line with the challenges identified, this thesis aims to contribute to the field of interaction design by offering a comprehensive framework for motion-based interaction design in large-volume spaces. These spaces, characterized by their size and complexity, present unique challenges in terms of seamless and intuitive user experiences. The framework presented bridges the gap between technical parameters and interpretable motion-based features, and is thus well-suited to a broad range of practitioners.