Human-Computer Interaction Using Deep Fusion Model-Based Facial Expression Recognition System

A deep fusion model is proposed for facial expression-based human-computer Interaction system.Initially,image preprocessing,i.e.,the extraction of the facial region from the input image is utilized.Thereafter,the extraction of more discriminative and distinctive deep learning features is achieved using extracted facial regions.To prevent overfitting,in-depth features of facial images are extracted and assigned to the proposed convolutional neural network(CNN)models.Various CNN models are then trained.Finally,the performance of each CNN model is fused to obtain the final decision for the seven basic classes of facial expressions,i.e.,fear,disgust,anger,surprise,sadness,happiness,neutral.For experimental purposes,three benchmark datasets,i.e.,SFEW,CK+,and KDEF are utilized.The performance of the proposed systemis compared with some state-of-the-artmethods concerning each dataset.Extensive performance analysis reveals that the proposed system outperforms the competitive methods in terms of various performance metrics.Finally,the proposed deep fusion model is being utilized to control a music player using the recognized emotions of the users.

An Optimal Scheme for WSN Based on Compressed Sensing

Wireless sensor networks(WSNs)is one of the renowned ad hoc network technology that has vast varieties of applications such as in computer networks,bio-medical engineering,agriculture,industry and many more.It has been used in the internet-of-things(IoTs)applications.A method for data collecting utilizing hybrid compressive sensing(CS)is developed in order to reduce the quantity of data transmission in the clustered sensor network and balance the network load.Candidate cluster head nodes are chosen first from each temporary cluster that is closest to the cluster centroid of the nodes,and then the cluster heads are selected in order based on the distance between the determined cluster head node and the undetermined candidate cluster head node.Then,each ordinary node joins the cluster that is nearest to it.The greedy CS is used to compress data transmission for nodes whose data transmission volume is greater than the threshold in a data transmission tree with the Sink node as the root node and linking all cluster head nodes.The simulation results demonstrate that when the compression ratio is set to ten,the data transfer volume is reduced by a factor of ten.When compared to clustering and SPT without CS,it is reduced by 75%and 65%,respectively.When compared to SPT with Hybrid CS and Clustering with hybrid CS,it is reduced by 35%and 20%,respectively.Clustering and SPT without CS are compared in terms of node data transfer volume standard deviation.SPT with Hybrid CS and clustering with Hybrid CS were both reduced by 62%and 80%,respectively.When compared to SPT with hybrid CS and clustering with hybrid CS,the latter two were reduced by 41%and 19%,respectively.