Food styling and food photography with generative AI

The objective of this study is to evaluate the aesthetic suitability of generative AI food images and to examine the potential role of AI in food styling and photography, including its strengths, weaknesses, opportunities, and threats. In this research, eight dishes from Turkish cuisine, Imambayıldı and Zeytinyağlı enginar (artichoke with extra virgin olive oil) for the olive oil theme, Adana kebab and Hünkâr beğendi for the main course theme, fırında sütlaç (baked rice pudding) and pumpkin dessert for the dessert theme, çay (Turkish tea) and Turkish coffee for the beverage theme, were produced separately using Adobe Firefly 3 and DALL-E 3 Artificial Intelligence (AI) applications. Real food photographs were also included for comparison. Thirty-one professional food stylists and photographers volunteered and participated in the study. Consequently, a total of 24 food images were created and evaluated by professionals according to six aesthetic criteria: lighting, color, composition, presentation, appropriateness of the props and background, and the creation of a mouth-watering sensation. The findings reveal no significant difference between the food photographs produced using the AI 1 application and real food photographs. Half of the images created by the AI 2 application also showed no significant differences compared to real images. However, significant differences were observed in five images between the two AI applications. Participants highlighted low costs, fast production, and flexibility as strengths of AI applications in food styling and photography. Conversely, weaknesses included the production of surreal images and aesthetic concerns. Opportunities were identified in fostering innovation, creativity, and new perspectives, while potential threats involved ethical and copyright concerns, overdependence on AI tools, and potential job displacement.