We proudly celebrate Amina Rakhimzhanova (아미나 락힘츠하노바)’s achievement, whose study, “감성 유사도 기반 레퍼런스 이미지 검색” (Reference Image Retrieval Based on Affective Similarity), received the Best Paper Award in the Poster Session at the Korean Society for Emotion and Sensibility. Her study investigates how computational methods can retrieve reference images that align with the nuanced affective and visual qualities expressed in designers’ concepts.

Amina developed a computational retrieval model that compares design concepts and reference images across eight emotional and visual factors: Mystique, Naturalness, Excitement, Smartness, Elegance, Thermal Perception, Correlated Color Temperature (CCT), and Color Complexity. The model was refined through parameter optimization and validated through evaluations with design experts, achieving an average retrieval success rate of 71.8%. The findings demonstrate the potential of affective similarity-based retrieval to support designers in discovering relevant visual inspiration and exploring ideas across diverse design domains. Congratulations to Amina on this meaningful achievement!

Abstract

Affective qualities are central to lighting design, yet methods for retrieving reference images using nuanced affective characteristics remain limited. This study presents a computational model that retrieves images matching design concepts through emotional and visual similarity across eight factors: Mystique, Naturalness, Excitement, Smartness, Elegance, Thermal Perception, Correlated Color Temperature (CCT), and Color Complexity. The model was refined through optimization using weighted parameters and validated through expert evaluations. The proposed approach achieved an average retrieval success rate of 71.8%, demonstrating its practical value for design workflows. These findings suggest that the method can support creative inspiration and reference image retrieval across diverse design domains.