This paper discusses the experimental details of speech enhancement using variational autoencoders (VAE). A joint VAE architecture is formulated, and a training protocol that strikes a balance between speech enhancement and VAE correctness is defined. Extended short-term objective intelligibility (ESTOI) is used to measure the intelligibility of enhanced speech. The proposed approach is implemented using MFCC and STFT features on a benchmark dataset and we report, on an average, 2 times improvement in ESTOI for enhanced speech using MFCC over STFT features across all noise levels. Further, the proposed approach using MFCC features shows significant improvement in denoising very noisy speech, as opposed to marginal improvement on relatively clean speech.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Speech Enhancement Using Variational Autoencoders


    Contributors:


    Publication date :

    2023-05-25


    Size :

    164195 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Deep Tracking Portfolios Using Autoencoders and Variational Autoencoders

    Urrego, Daniel Aragón / Nieto, Oscar Eduardo Reyes / Quimbayo, Carlos Andrés Zapata | Springer Verlag | 2024


    Variational Autoencoders

    Ghojogh, Benyamin / Crowley, Mark / Karray, Fakhri et al. | Springer Verlag | 2022


    Certifiably Robust Variational Autoencoders

    Barrett, Ben / Camuto, Alexander / Willetts, Matthew et al. | ArXiv | 2021

    Free access

    Mixed-curvature Variational Autoencoders

    Skopek, Ondrej / Ganea, Octavian-Eugen / Bécigneul, Gary | ArXiv | 2019

    Free access