Bees Local Phase Quantisation Feature Selection for RGB-D Facial Expression Recognition

Feature selection can be defined as an optimisation problem and solved by bioinspired algorithms. The Bees Algorithm (BA) returns great performance in the feature selection optimisation task. On the other hand, local phase quantisation (LPQ) is a frequency domain feature that has excellent performance on depth images. Here, after extracting LPQ features from RGB (colour) and depth images from the Iranian Kinect Face Database (IKFDB), the Bees feature selection algorithm is applied to select the desired number of features for final classification tasks. IKFDB is recorded with Kinect sensor V.2 and contains colour and depth images for facial and facial microexpression recognition purposes. Here, five facial expressions, Anger, Joy, Surprise, Disgust and Fear, are used for final validation. The proposed Bees LPQ method is compared with Particle Swarm Optimisation (PSO) LPQ, PCA LPQ, Lasso LPQ, and just LPQ features for classification tasks with Support Vector Machines (SVM), K-Nearest Neighbourhood (KNN), Shallow Neural Network and Ensemble Subspace KNN. The returned results show a promising performance of the proposed algorithm (99% accuracy) in comparison with others.