A Novel Mobile Application for Personalized Hair Care Using Advanced Machine Learning Techniques

Recent advancements in computer vision and machine learning have enabled personalized hair care, advancing beyond traditional methods by catering to individual hair types and health. This paper presents Glowllp, a mobile application that provides personalized hair care recommendations through three primary modules: image quality assessment, hair type detection, and health prediction. The image quality assessment module employs edge detection and SVM classifiers to filter blurry images, achieving over 95% accuracy. For hair type classification, a hybrid model of Mask R-CNN and Vision Transformers (ViTs) identifies hair types with 92% accuracy. The health assessment module predicts hair loss and other conditions using Random Forest and XGBoost models with an F1 score of 0.91, while K-means clustering further tailor's recommendations. By integrating these modules, Glow Up delivers a comprehensive, user-friendly solution that enhances hair care personalization.