Face Embedding

Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation
Elegant ways to check the so-called “Manifold Hypothesis” for FaceID using geoemtric (Ninkiwsky-based) and probabilistic approaches.
Metric Learning for facial descriptors
Problem of metric learning applied to face recognition.
Neural Nets recent success and introduction to FaceID
Overview of advances in Neural Nests, Deep Learning, and Introduction to FaceID.