Development of Alexa Voice Services Software Development Kit For Speech Recognition Engine For Internet of Things

Abstract

In 1990s, Internet connectivity began. Internet established a link between people through PC. With the growth in area of Electronics, Communication, Information Technology and Computer Science in terms of technology, reliability, efficiency and availability IoT also started growing rapidly. Cisco predicts that the total number of connected devices in IoT will reach up to 50 Billion (approx.) by year 2020. Google Home, Amazon Echo, Apple Siri have been added to the journey of IoT with the evolution of new technology and algorithms in a decade and this journey will continue over many decades. Speech Recognition provides a capability to user to interact with device by Speech Interface. People with hearing disability can use speech recognition to convert a telephone caller’s speech into text format. In future, advancement and expansion of Speech Recognition techniques and Artificial Intelligence, Nueral Networks, Echo Cancellation, Beam Forming and Noise Reduction techniques will enhance the quality and efficiency of Speech Recognition Engines. Finally result shows that by improving some key features and parameters such as Wake Word False Rejection Rate, Wake Word Detection Delay, Wake Word Detection Delay, Multi Room Music (MRM) playback, Response Accuracy Rate and Audio Voice Quality of Speech Recognition Engine better user experience can be achieved.