Haruyuki Fujii

Generative AI is increasingly applied in architectural research, from automated ideation and reshaping design workflows to design education. Despite the increasing realism of synthetic imagery, several research gaps remain including alignment, plausibility, explainability, and control. This study focuses on alignment with human perceptions, specifically examining how synthetic architectural façade imagery aligns with geographical, objective, and affective text descriptors. We propose a pipeline that applies a Latent Diffusion Model to generate façade images and then evaluate this alignment through both AI-based and human-based evaluations. The results reveal that while images generated with geographical prompts are notably aligned, they also show clear biases. The results also reveal that images synthesised from objective descriptors (e.g., angular/curvy) are more aligned with human perceptions than affective descriptors (e.g., utopian/dystopian). These initial results highlight the opportunities and limits of current generative AI models, hinting at data biases and the potential lack of embodied understanding to grasp the complexity in experiencing architecture. Limitations of the study remain. Future work can expand on exploring cultural biases and semantic overlaps, and in testing more advanced embodied AI models and methods.

Generative AI is increasingly applied in architectural research, from automated ideation and reshaping design workflows to design education. Despite the increasing realism of synthetic imagery, several research gaps remain including alignment, plausibility, explainability, and control. This study focuses on alignment with human perceptions, specifically examining how synthetic architectural façade imagery aligns with geographical, objective, and affective text descriptors. We propose a pipeline that applies a Latent-Diffusion-Model to generate façade images and then evaluate this alignment through both AI-based and human-based evaluations. The results reveal that while images generated with geolocation prompts are notably aligned, they also show regional biases. The results also reveal that images synthesized from objective descriptors (e.g., angular/curvy) are more aligned with human perceptions than affective descriptors (e.g., utopian/dystopian). These initial results highlight the opportunities and limits of current Generative AI models, hinting at data biases and the potential lack of embodiment/experience/memory of these models to grasp the complexity in experiencing architecture.

Gaze-sharing-the act of sharing gaze information among several people-is increasingly valued by HCI researchers. Past studies have demonstrated its benefits for education, especially when learning to cope with well-defined problems. However, since design often deals with ill-defined problems, further research is needed to explore its potential benefits for design education. One critical activity in design education is that of design critique. This study held controlled conversation experiments during a design critique under three conditions: without technological facilitation (natural), with real-time view-sharing, and with real-time gaze sharing (RTGS). It was found out that an RTGS system can enhance communication in design education in two aspects: 1) by reducing moments of silence, an RTGS system can streamline communication in studios (critiques facilitated by RTGS were the most efficient among all three); 2) the instructor and the students are found to benefit more in the gaze-sharing condition, as it eases access to visual information and encourages discussion in the studio, compared with studios in other two conditions. These findings suggest great potential for integrating RTGS into future studio practices for enhancing communication in design education.

This study aims to understand how the characteristics of the architectural space are manifested by applying two inquiry methods, the visibility graph analysis (VGA) and the analysis of occupants' actions in combination. The visibility properties of a house are analysed in relation to the room layout and the distribution of occupant actions in two conditions, with or without an exterior. Subsequently, the sociality of vernacular houses is explored by applying these research methods together. This study uses five typical traditional private houses in Izena, Ryukyu as a case study.

The study of interpretation is of major importance for our understanding of design cognition. When interacting with design representations, designers often rely on metaphorical descriptions as interpretive devices, which aid in coping with the task at hand. Consequently, such descriptions can enlighten us regarding the designer's perspective of the situation, and their analysis can deepen our knowledge of design cognition. We observe designers as they metaphorically interpret design representations during a simple task of spatial configuration, and introduce an approach for modeling this practice, as a means for getting insights into the designer's mental world. In this, we draw on traditional practices of protocol analysis, as well as on state-of-the-art theoretical frameworks for situated design and discourse analysis. Our integrated approach demonstrates how important relations between external and internal reality in design activity can be mapped and visualized. This sheds some light on the cognitive process of interpretation in design. The proposed method can both serve as a basis for detailed analyses of design cognition and for the enhancement of current models for situated design agents.

Metaphors are powerful tools for design, enabling designers to encapsulate sets of properties and relations as short verbal descriptions. This paper aims to clarify how simple spatial configurations may emerge from concise metaphoric descriptions at the conceptual design phase. To this aim, we propose a framework for a metaphor-based design process. As a basis for the framework, we introduce the concept of “complementary visual potential” – a property which ties the spatial configuration of the objects in the composition with their metaphoric roles. The framework is developed by studying the practice of metaphor-based spatial configuration design in Japanese rock gardens. Accordingly, it is implemented and tested in this context by attempting to generate alternative designs for an existing rock composition in the famous garden of Ryōan-ji. This is followed by a discussion of its possible implications and potential for generalization to other areas of design.