Saved in:
Bibliographic Details
Main Authors: Pegg, Samuel, Li, Kai, Hu, Xiaolin
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2309.17189
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913275416936448
author Pegg, Samuel
Li, Kai
Hu, Xiaolin
author_facet Pegg, Samuel
Li, Kai
Hu, Xiaolin
contents Audio-visual speech separation methods aim to integrate different modalities to generate high-quality separated speech, thereby enhancing the performance of downstream tasks such as speech recognition. Most existing state-of-the-art (SOTA) models operate in the time domain. However, their overly simplistic approach to modeling acoustic features often necessitates larger and more computationally intensive models in order to achieve SOTA performance. In this paper, we present a novel time-frequency domain audio-visual speech separation method: Recurrent Time-Frequency Separation Network (RTFS-Net), which applies its algorithms on the complex time-frequency bins yielded by the Short-Time Fourier Transform. We model and capture the time and frequency dimensions of the audio independently using a multi-layered RNN along each dimension. Furthermore, we introduce a unique attention-based fusion technique for the efficient integration of audio and visual information, and a new mask separation approach that takes advantage of the intrinsic spectral nature of the acoustic features for a clearer separation. RTFS-Net outperforms the prior SOTA method in both inference speed and separation quality while reducing the number of parameters by 90% and MACs by 83%. This is the first time-frequency domain audio-visual speech separation method to outperform all contemporary time-domain counterparts.
format Preprint
id arxiv_https___arxiv_org_abs_2309_17189
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle RTFS-Net: Recurrent Time-Frequency Modelling for Efficient Audio-Visual Speech Separation
Pegg, Samuel
Li, Kai
Hu, Xiaolin
Sound
Computer Vision and Pattern Recognition
Audio and Speech Processing
Audio-visual speech separation methods aim to integrate different modalities to generate high-quality separated speech, thereby enhancing the performance of downstream tasks such as speech recognition. Most existing state-of-the-art (SOTA) models operate in the time domain. However, their overly simplistic approach to modeling acoustic features often necessitates larger and more computationally intensive models in order to achieve SOTA performance. In this paper, we present a novel time-frequency domain audio-visual speech separation method: Recurrent Time-Frequency Separation Network (RTFS-Net), which applies its algorithms on the complex time-frequency bins yielded by the Short-Time Fourier Transform. We model and capture the time and frequency dimensions of the audio independently using a multi-layered RNN along each dimension. Furthermore, we introduce a unique attention-based fusion technique for the efficient integration of audio and visual information, and a new mask separation approach that takes advantage of the intrinsic spectral nature of the acoustic features for a clearer separation. RTFS-Net outperforms the prior SOTA method in both inference speed and separation quality while reducing the number of parameters by 90% and MACs by 83%. This is the first time-frequency domain audio-visual speech separation method to outperform all contemporary time-domain counterparts.
title RTFS-Net: Recurrent Time-Frequency Modelling for Efficient Audio-Visual Speech Separation
topic Sound
Computer Vision and Pattern Recognition
Audio and Speech Processing
url https://arxiv.org/abs/2309.17189