31 March 2024 in Journal of Computational Design
This article aims to explore the integration of artificial intelligence (AI) as a design tool in interior design education. The research examines the students' interior design studio project outcomes over the usage of AI in creating conceptual images, and the implementation of the AI-created concept to the overall space. In the research, students' projects are divided into two groups of 5 according to sufficient or insufficient prompts for the "AI generated" conceptual images. Barnard's (1992) CAIDC (Consensual Assessment of Interior Design Creativity) scale was used for the assessment. Mann-Whitney U Test was conducted for the results. We understand that there is no significant difference between writing sufficient or insufficient prompts in the concept development phase of interior design projects according to the Barnard (1992)’s design merits. It has been confirmed that the main factor that influences this regard is the need for an appropriate "concept analysis" to adapt the concept generated with AI to the specified project spaces.
15 December 2023 in Interaction Design and Architecture(s)
2 September 2023 in Interiors
27 June 2024 in Mobilya ve Ahşap Malzeme Araştırmaları Dergisi
This study at Antalya Bilim University's Interior Architecture and Environmental Design Department explores an innovative educational model that combines the "Human Factors in Interior Space" and "Furniture Design" courses. Aimed at third-year students, this approach bridges theoretical understanding with practical application, focusing on ergonomic design principles to create functional, human-centric furniture. Thus, it allows students to combine the concepts of anthropometry and ergonomics. Through a two-phased methodology, students first engage with traditional anthropometric methods to construct a standard-sized seating element, then advance to digital techniques, including photogrammetry, to design a custom cardboard seating element tailored to individual needs. The incorporation of digital tools such as Agisoft and Autodesk ReCap, alongside traditional plaster molding, facilitates a comprehensive understanding of ergonomic design. The study underscores the importance of integrating manual and digital methodologies in interior design education, enhancing students' ability to innovate and adapt to modern design challenges.
1 January 2025 in PLANARCH - Design and Planning Research
This study addresses the challenge of evaluating how effectively artificial intelligence (AI) models can interpret and visualize basic design principles and elements—a skill essential in creative disciplines such as art and design education. The core aim is to examine whether AI-generated visual outputs can reflect a deliberate understanding of design concepts, particularly within abstract symmetrical and asymmetrical compositions. To explore this, the study analyzes the performance of 18 widely used text-to-image AI models, based on five design principles—rhythm, movement, contrast, emphasis, and balance—and five design elements—dot, line, shape, color, and texture. Each model was prompted to generate compositions aligned with these criteria, and the outputs were evaluated based on three primary criteria: understanding the design elements, understanding the design principles, and understanding the balance. Scores ranged from 1 (least accurate) to 5 (most accurate), and averages were calculated for comparative analysis. The results revealed that while models like Microsoft Designer and DALL-E occasionally produced strong outputs, they lacked consistency across different balance types. In contrast, models such as Craiyon and Flux.ai consistently underperformed. Notably, high scores in asymmetrical balance often inflated overall averages, masking deficiencies in design comprehension. The study concludes that symmetrical balance serves as a more reliable indicator of AI proficiency and emphasizes the need for holistic evaluation frameworks. These findings offer valuable insights for integrating AI tools into design education and assessing their pedagogical potential.
1 October 2025 in Turkish Online Journal of Design Art and Communication
This study evaluates the performance of three widely-used text-to-image AI models—DALL-E, MidJourney, and Stable Diffusion—in generating interior design visuals. Five standardized prompts, covering essential design elements such as architectural features, style, furniture, lighting, textiles, accessories, and layout, were provided to each AI model. The resulting visuals were then assessed by 15 experienced professionals and academics with over 10 years of expertise in interior design. Using a detailed rating system, the study found that MidJourney consistently outperformed the others, excelling in categories like design style, textiles, and lighting. DALL-E proved effective in generating architectural features but struggled with finer details, particularly in textiles and lighting. Stable Diffusion lagged behind in overall performance, especially in categories like layout and furniture. The study highlights the importance of prompt specificity in shaping AI-generated visuals and suggests areas for future improvement in AI models, particularly in complex design tasks.