Multi Modal Aptitude Detection System for Smart Office

Abstract

Smart offices are places that aim to provide a favorable, interactive, and healthy environment for the employees. As a result, their working efficiency is enhanced. In order to achieve these numerous systems such as lie detector, emotion detection, and recognition, automation of office equipment has been proposed and implemented. To contribute toward one of the goals of smart offices, a novel ”multi-modal aptitude detection system” is being proposed. The proposed system helps in analyzing the working aptitude of employees in order to help them in providing the more conducive and healthy working environment. In this paper, a detailed layered framework, implementation details, and achieved outcomes related to the proposed novel system are presented. The results achieved show that this novel system after successfully becoming a part of smart office can bring significant enhancements in the efficiency of employees. The proposed system helps in learning and identifying job preferences of employees based on their working aptitude. The system is validated and showed an accuracy of 96% in determining an employee's aptitude. The results are further verified by computing the F1 scores, and more details regarding testing and validation are provided.

Keywords

• Wireless sensor networks
• wireless sensor based applications
• classification
• machine learning
• signal processing
• neural networks