サリエンシーディテクションを用いた商品陳列における注目推定およびトップダウン型注意が及ぼす影響

抄録

In the marketing field, visual information significantly influences the purchasing behavior of consumers. This study focuses on the relationship between product display and visual information, aiming to provide an objective index of attention attractiveness that can guide layout decisions and improve package designs. We propose a system for estimating the area of interest (AOI) in a product display using saliency detection, a typical method reflecting bottom-up attention. The estimation system uses images captured in an actual store. Employing a saliency detection method based on differences in spatial frequency spectra, the system calculates the saliency of each AOI for each product area and outputs it as an estimate of the AOI . We experimentally measured the eye gaze of 12 subjects to assess the adaptability of the system and the impact of top-down attention. Subsequently, we compared the estimated system with the actual gaze distribution. The results showed no overall correlation; however, a positive correlation was observed among the experimental collaborators, suggesting the potential for estimation. Furthermore, a positive correlation was noted in many non-planned purchases; however, this correlation decreased in planned purchases. These findings indicate that increased top-down attention results in an attention pattern different from that of the estimation map.