Computational model for predicting user aesthetic preference for GUI using DCNNs

Visual aesthetics is vital in determining the usability of graphical user interface (GUI). It can strengthen competitiveness interactive online applications. Human aesthetic preferences for GUI are implicit and linked to various aspects perception. In this study, an image database was constructed with 38,423 design works collected from Huaban.com, a popular social network website art sharing, collection, exhibition China. The numbers collection likes each work were used as annotation represent preference levels. Deep convolutional neural networks applied evaluate GUIs, based on large dataset images ground-truth annotations. experimental result indicated feasibility proposed method, mean squared error (MSE) 0.0222 prediction MSE 0.0644 best model performance Squeeze-and-Excitation-VGG19 (SE-VGG19). This study aims build database, explore practical objective evaluation aesthetics.