Untitled Item

This study explores the emotional perception of AI-generated architectural images through an in-depth comparison between two groups of participants: architecture students (Group A, n=389) and non-architecture students (Group B, n=400). We revealed the complexity of emotional expression, especially in conveying six basic emotions. Although the expression of joy was the most successful, the conveyance of other emotions posed challenges. We also discovered significant differences in emotional perception between the two groups, particularly in joy and surprise, which may be related to professional training and educational background. Additionally, the study emphasizes the importance of architectural images, particularly the emotion of happiness, in emotional response, and unveils a profound connection with human emotions. We also found that AI-generated indoor images have a better capacity for emotional rendering than outdoor images, possibly reflecting AI's challenges in capturing emotional elements in complex environments. Moreover, binary emotional analysis revealed that AI might be more effective when dealing with a broader category of emotions. The findings of this study not only highlight the challenges and potentials of AI in emotional expression but also illustrate how professional backgrounds shape perception and cognitive processes, providing new directions and challenges for future research.