Development of Multimodal emotion classification system with usability features

Abstract

Evolution of technologies leads to the development of various newlineapplications which enhances the efficacy of the human-kind. Applications newlinethat are being developed should meet its requirement with respect to the newlinefunctionalities and also their accessibility. Usability of an application newlinerefers the quality of the accessibility that the application provides. newlineRecently, usability engineering gained attention because of system factors newlinesuch as, people s satisfaction, competitors in the market and so on. newlineThe conventional way of estimating the usability metrics of an newlineapplication is by manually getting the input from the test users and some newlineof the selected real time users through interruptions. Such interruptions do newlinehave a drawback of changing the mood of the users for particular period newlineof time. Also, receiving the frequent feedback is also not possible in newlinereal time scenario. For addressing this issues, the research work in this newlinethesis proposes a novel system for estimating the usability metrics without newlinemanual interruptions using multi-modal emotional features from actions newlineof humans. newlinevi newlineA research work for designing the game for enhancing memory newlinepower of kids is proposed and also effective process of designing kiosk newlinesystems which are always act as an interface for the people. Another newlineproposed work in the thesis addresses the estimating the emotion of the newlineuser from the facial feature of the users. Multimodal features from humans newlinesuch as facial image, action log and voice, based usability estimation newlinesystem is also proposed. newline newline