Description

Speech Recognition using Multiscale Scattering of Audio Signals and Long Short-Term Memory of Neural Networks

Communication is the one of the key elements of interaction. Humans used different languages to communicate with one another whereas technology-based devices use their own language to process its‟ commands. In order to understand the audio language used by humans, machines use different techniques to convert speech to machine readable form called speech recognition. There have been several machine learning and deep learning techniques that are used to recognize audio signals. This paper takes one of the most classic examples of the speech recognition domain, the spoken digits recognition and talks about a technique called wavelet scattering to initially extract useful information from the signals. This information is further sent to a Long Short-Term Memory (LSTM) network to classify the signals. Keywords – Deep Learning, Long Short-Term Memory (LSTM), Multi-Scale Scattering, Neural Networks, Speech Recognition