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