Deep Learning Framework for Exploring Customized CNN with Customer Demographics and Salon Visuals

This research explores the integration of visual aesthetics and demographic data to enhance customer satisfaction in the beauty salon industry through a custom Convolutions Neural Network (CNN) model. Traditional salon experiences rely on visual elements such as interior design, ambiance, and branding, which shape customer perceptions. However, the influence of demographic factors such as age, gender, and income on these visual cues has been under explored. To address this gap, a CNN model was developed to combine visual and demographic inputs, enabling the prediction of customer preferences and segmentation. By analyzing images of salon environments and integrating customer data, the model offers actionable insights for salon owners to optimize design and services. The findings show that the custom CNN, enhanced with self-attention mechanisms, achieved high accuracy in predicting customer preferences, with a training accuracy of 98.1 % and a validation accuracy of 98.7%. This approach has practical applications in improving salon strategies and customer satisfaction.