Synthesis, characterization and applications of silver(I), gold(I) and palladium(II) complexes derived from coumarin tethered N–heterocyclic carbenes

Abstract

Over the last few years, facial expression recognition has obtained attention from scientists in psychology, computer science, medicine and related fields, which has an extensive range of applications in the field of social interaction, social intelligence, autism detection and Human-computer interaction. newline newlineAlthough human beings can identify the facial expressions effortlessly, reliable automatic facial expression recognition by machines is still a challenge. The primary purpose of this research is to develop a robust framework for the automatic classification of facial expressions and cognitive states from images and videos. newline newlineThe proposed facial expression recognition system intends to analyze the basic facial expressions and cognitive states from image and videos by extracting significant features of a face. Seven distinct primary facial expressions such as happy, disgust, surprise, anger, sad, fear and neutral have been used for the experimental purpose. Also, the research work proposed to recognize the cognitive states such as interested, bored, thinking, happy and unsure by combining facial expressions and hand-over-face gestures. newline newlineIn this research, a novel framework is presented to recognize the facial expressions, which enhances the efficiency and speed of recognition system by extracting significant features of a face. In the proposed framework, feature representation and extraction are done by using Median Robust Extended Local Binary Patterns (MRELBP) based Histogram of Oriented Gradients (HOG). Later, the dimensionalities of the obtained features are reduced using Compressive Sensing (CS) algorithm and classified using multiclass SVM classifier. We investigated the performance of the proposed framework on two public databases such as CK+ and JAFFE data sets. Using the proposed hybrid approach, we achieved an average accuracy of 93% and above on JAFFE and CK+ databases. The investigational results prove that the proposed framework is robust for recognizing facial expressions with varying illuminations and poses. newline newlineHowever, most research focuses on posed expressions, near frontal recordings and they consider hand occlusions as noise. Especially, in near frontals like people sitting in front of the desktop/laptop, frequently hold their hands close to the face which occludes the face and limits the accuracy of facial expression recognition. There is experimental proof that some of the hand-over-face poses can be used for the detection of cognitive states. In our research, we propose to use hand-over-face gesture as novel cues and integrate facial expressions with hand-over-face poses for the identification of cognitive states like interested, bored, unsure, happy and thinking. The proposed system is robust to changes in facial expressions, hand action, occlusions and performs an average recognition accuracy of 90.51%. newline newline