FHP: Facial and Hair Feature Processor for Hairstyle Recommendation

Abstract

Hairstyles contribute majorly to an individual’s physical appearance. It is commonly observed that there is a huge gap between customer’s expectations and the barber’s understanding leading to post-haircut dissatisfaction. A major cause of this problem is that users lack the understanding of utilizing feature information while selecting a suitable hairstyle. A model that extracts essential features for decision-making can improve customer satisfaction. In this paper, we propose the FHP architecture that extracts features from user-submitted images that are crucial to hairstyle recommendations. The architecture consists of two major pipelines for extracting facial and hair features. The process starts with image segmentation and parsing. The segmented facial and hair regions are given as input to the two pipelines for extracting face shape, skin tone, hair texture, and hair color. These four features can then be used in the development of an expert system for hairstyle recommendations.