Development of a System for Human Language Commands and Control for a Quadcopter Application

Abstract

This project intends to control and navigate the Quadcopter in an undefined nature using sensors and using only voice recognition. This project aims to improve speech recognition software system and test it. The system processes human voice input and answers the queries of the user.

This system is planned task related with the action of the human voice. In this sense, the application fields can change from speech production, storage, transmission and recognition processes. 

Hence, this thesis uses this technology for speech recognition. The main emphasis is given to the Hidden Markov Models, Artificial Neural Network, and Linear Predictive Coding. The objective of this project is to let people have the experience of free flight being independent of a map and without using a mobile phone. This project works using a computer only. My objective was to achieve alternative result in contrast to different methods/systems.

This application is tested with speakers where the test included different tones, emotions and lengths of sample for related methodology. The reports of the system data result are used in the rating and analysis of obtained results. Due to this, test shows us that test the improved speech recognition PC system for mankind language, which was tested on various feelings and lengths is a more accurate system with less error than former researches.