As described in the previous chapter, the selection and localization of facial landmarks are crucial for subsequent tasks such as image crop, image alignment, and geometric feature extraction. In this chapter, we introduce a new category of geometric descriptors, angle features, which capture the relative orientation between landmarks and provide complementary information beyond distances and ratios. We also review common types of facial landmarks and discuss representative localization algorithms that balance accuracy, robustness, and computational efficiency. Building upon these foundations, we present a genetic algorithm-based feature selection strategy designed to identify the most informative and discriminative landmarks for facial beauty analysis. Finally, we evaluate and compare the performance of different landmark models and their corresponding feature sets in facial beauty prediction tasks, and conduct a detailed ablation study to analyze the contribution of each feature subset to overall prediction accuracy.