Artificial narrow intelligence adaptive audio processing
Authors
Singh, Sahib Preet
Issue Date
2019
Degree
MSc in Data Analytics
Publisher
Dublin Business School
Rights holder
Rights
Items in eSource are protected by copyright. Previously published items are made available in accordance with the copyright policy of the publisher/copyright holder.
Abstract
The age of AI is engulfing us with experts attempting to envision a future driven by the rise of this far-reaching technology. Significant progress has been made during 2018-2019 for AI and deep learning in particular. The rise of ANI ( Artificial Narrow Intelligence) for image and video processing has emerged alongside advances in the key domains of language, control, and decision-making. Following on from the successes in image and video processing, research has extended to the domain of audio processing. This paper outlines impressive technological advancements made in ANI enabled audio processing. The research includes a practical review of the processes involved in building audio processing AI systems. The paper describes a method for feature extraction, modelling and deployment of a voice recognition model on the Raspberry Pi3 , the smartest, most robust and smallest computer circuit known to mankind. A visual analysis of the extracted features is performed. Performance parameters for audio processing are balanced for the layers of a convolutional neural network. The research results demonstrate a model accuracy of 98.56%, an improvement on the accuracy reported by previous research.