Florence 2D/3D Hybrid Face Dataset

bridges the gap between 2D, appearance-based recognition techniques, and fully 3D approaches (Bagdanov, Del Bimbo, and Masi)