• Lalita Devi
• Er. Parmeet Kaur

In this work, it presents an improved universal codebook-based speech enhancement framework that relies on a single codebook to encode both speech and noise components. The aim of the noise reduction algorithms is to estimate the clean speech signal from the noisy recordings in order to improve the quality and intelligibility of the enhanced signal. Given the good performance of deep learning in signal representation, a deep auto encoder will be employed in this work for accurately modelling the clean speech spectrum. Due to this, it proposes a method for speech enhancement using codebook based method iteratively. In this work, it provides speech enhancement under different noisy conditions. The speech power spectrum will be measured greatly for different types of speech sound. It shows the performance comparison of proposed method with actual methods and proposed method shows better improvement in ∆ILD, ∆IPD values.

Speech Processing, Speech Enhancement, etc.

"Improvement in Speech Quality by Modified Codebook Based Technique", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 10, page no.1547-1552, October-2020, Available :http://www.jetir.org/papers/JETIR2010199.pdf

"Improvement in Speech Quality by Modified Codebook Based Technique", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 10, page no. pp1547-1552, October-2020, Available at : http://www.jetir.org/papers/JETIR2010199.pdf