Speech Enhancement Aided For Voice Activation and Denoising Using CNN Algorithm

Description

This paper aims to address the problem of audio denoising by developing a deep convolutional neural network (CNN) model. The model is trained on a dataset consisting of clean audio signals and corresponding noisy signals augmented with common environmental noises. By leveraging the capabilities of deep CNNs, the model aims to extract the clean audio signal and suppress the noise effectively. The trained model can be deployed on either the client or server side for denoising purposes. The project’s goal is to provide a reliable and efficient solution for enhancing audio quality in diverse applications.