Automatic Speech Recognition and its Visual Perception Via a Cymatics Based Display

The objective of this work is to provide an opportunity for people with impaired hearing to appreciate the auditory art. In this context, the science of Cymatics is employed for a systematic display of auditory messages and commands. The physical phenomenon of Cymatics allows the visualization of sounds by applying the basic principles of wave mechanics. The effects of sound waves can be shaped on an appropriate media. In this framework, an innovative method is devised. It allows a systematic visualization of speech. This work deals with the design and development of three main modules. The first module identifies the message from a speech by employing the speech processing and machine learning based automatic speech recognition (ASR) method. The second module transforms the identified message into a methodical excitation. The third module provides a systematic representation of this excitation wave through the physical phenomenon of Cymatics. The proposed system prototype is realized. The system functionality is verified with the help of an experimental setup. An average isolated word recognition accuracy of 97% is achieved. Results prove the concept and working principle of the devised system.