Fusing Multiple Bandwidth Spectrograms for Improving Speech Enhancement

Shi, Hao; Shu, Yuchun; Wang, Longbiao; Dang, Jianwu

doi:10.23919/apsipaasc55919.2022.9980180

Cited by 2 publications

(1 citation statement)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The findings highlight the superiority of U-Net models in the time-frequency complex domain, achieving substantial improvements in speech enhancement measures across low Signal-to-Noise ratio scenarios. In [17], a based convolutional system is developed to enhance spectral information by simultaneously utilizing multiple bandwidth spectrograms, specifically augmenting wider bandwidth (16 ms and 8 ms) spectrograms as auxiliary information. Experimental results on the VB dataset demonstrate that incorporating different bandwidth spectrograms provides supplementary information, resulting in an over 0.1 improvement, with the embedding dimension influencing the fusion strategy in the encoder.…”

Section: State Of the Artmentioning

confidence: 99%

Analyzing the Influence of Diverse Background Noises on Voice Transmission: A Deep Learning Approach to Noise Suppression

Nogales,

Caracuel-Cayuela,

García-Tejedor

2024

Applied Sciences

View full text Add to dashboard Cite

This paper presents an approach to enhancing the clarity and intelligibility of speech in digital communications compromised by various background noises. Utilizing deep learning techniques, specifically a Variational Autoencoder (VAE) with 2D convolutional filters, we aim to suppress background noise in audio signals. Our method focuses on four simulated environmental noise scenarios: storms, wind, traffic, and aircraft. The training dataset has been obtained from public sources (TED-LIUM 3 dataset, which includes audio recordings from the popular TED-TALK series) combined with these background noises. The audio signals were transformed into 2D power spectrograms, upon which our VAE model was trained to filter out the noise and reconstruct clean audio. Our results demonstrate that the model outperforms existing state-of-the-art solutions in noise suppression. Although differences in noise types were observed, it was challenging to definitively conclude which background noise most adversely affects speech quality. The results have been assessed with objective (mathematical metrics) and subjective (listening to a set of audios by humans) methods. Notably, wind noise showed the smallest deviation between the noisy and cleaned audio, perceived subjectively as the most improved scenario. Future work should involve refining the phase calculation of the cleaned audio and creating a more balanced dataset to minimize differences in audio quality across scenarios. Additionally, practical applications of the model in real-time streaming audio are envisaged. This research contributes significantly to the field of audio signal processing by offering a deep learning solution tailored to various noise conditions, enhancing digital communication quality.

show abstract

Section: State Of the Artmentioning

confidence: 99%