Abstract

Brand identity (BI) serves as the embodiment of a company's values and characteristics. The study discussed in this session develops a machine-learning framework to quantify the effects of visual design elements on brand identity (BI) in packaging. Using VGG16 and Grad-CAM, the research analyzes key features like logos, shapes, and contrast, providing objective metrics for brand recognition. The findings highlight how machine learning can assist in maintaining BI consistency while adapting to market trends, offering brand managers data-driven tools to optimize packaging design for improved consumer perceptions and loyalty.