Estudo dos Avanços na IHM: a integração de Modelos de Visão-Linguagem e ConceptGraphs

Abstract

Human-machine interaction (HMI) has evolved significantly with the advancement of Vision-Language Models (VLMS) and ConceptGraphs. VLMS combine advanced text and image processing capabilities, enabling machines to understand and respond to visual and linguistic information in an integrated manner. On the other hand, ConceptGraphs are frameworks that capture complex semantic elationships in three-dimensional environments, facilitating interpretation and planning based on spatial contexts. This work explores and studies how the integration of VLMS and ConceptGraphs is redefining human machine interaction in diverse contexts. The theoretical foundation discusses the principles and applications of these technologies, highlighting their architectures, training and practical applications in advanced virtual assistants, intelligent user interfaces and other forms of human-computer interaction. The literature review analyzes relevant case studies and examples, providing insights into observed benefits, limitations faced, and opportunities for future research. The critical analysis and evaluation of the results discusses the impact of these technologies on efficiency, accessibility and user experience, while the conclusion summarizes key learnings and suggests directions for continued HMI development with VLMS and ConceptGraphs.