This document discusses rotation invariant face recognition using three feature extraction techniques: Rotated Local Binary Pattern (RLBP), Local Phase Quantization (LPQ), and Contourlet transform. It first extracts features from input face images using these three techniques. It then applies Linear Discriminant Analysis to reduce the feature dimensions. Finally, it uses k-Nearest Neighbors classification to perform face recognition on the Jaffe dataset. Experimental results show that the face recognition accuracy without LDA is 99.06% and increases to 100% when LDA is used for feature dimension reduction.