Rehaussement de la parole basée sur l’apprentissage profond (Deep Learning)

Abstract

In this project, we proposed the use of a pre-trained model based on Deep Learning to reduce noise in an audio signal. For this, we used a particular environment, named "Colaboratory", often shortened to "Colab". This latter is suitable for Machine Learning and data analysis. The model used allowed us to obtain noise-free speech signals. In this study, we also explored the theoretical concepts starting with the major importance of Deep Learning, followed by the explanation of the basics of Artificial Intelligence, Machine Learning and Deep Learning, as well as their evolution over time decades.