Abstract

In immersive virtual experiences, it is often necessary to select appropriate emotion recognition tools; however, different tools differ significantly in operational stability, response characteristics, and contextual adaptability, and current selection practices mostly rely on experience rather than structured methods. Under unified immersive experimental conditions, this study conducted multi-tool testing of visual-behavior, speech emotion recognition, and physiological-signal tools, and analyzed their performance differences over the complete experience process and across different interaction contexts. The results show that the various tool categories exhibit a stable differentiated structure in continuous operational capability and change-response characteristics, together with clear context dependence. On this basis, a two-dimensional applicability structure model is proposed, and a hierarchical decision-making framework for emotion recognition tool selection is constructed to realize a structured mapping from evaluation objectives and interaction conditions to tool types. Verification results show that the framework recommendations are consistent with the experimental performance structure, indicating basic feasibility.

Download

Share

COinS