MLCA-AVSR: Multi-Layer Cross Attention Fusion based Audio-Visual Speech Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, He, Guo, Pengcheng, Zhou, Pan, Xie, Lei
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913304170987520
author Wang, He
Guo, Pengcheng
Zhou, Pan
Xie, Lei
author_facet Wang, He
Guo, Pengcheng
Zhou, Pan
Xie, Lei
contents While automatic speech recognition (ASR) systems degrade significantly in noisy environments, audio-visual speech recognition (AVSR) systems aim to complement the audio stream with noise-invariant visual cues and improve the system's robustness. However, current studies mainly focus on fusing the well-learned modality features, like the output of modality-specific encoders, without considering the contextual relationship during the modality feature learning. In this study, we propose a multi-layer cross-attention fusion based AVSR (MLCA-AVSR) approach that promotes representation learning of each modality by fusing them at different levels of audio/visual encoders. Experimental results on the MISP2022-AVSR Challenge dataset show the efficacy of our proposed system, achieving a concatenated minimum permutation character error rate (cpCER) of 30.57% on the Eval set and yielding up to 3.17% relative improvement compared with our previous system which ranked the second place in the challenge. Following the fusion of multiple systems, our proposed approach surpasses the first-place system, establishing a new SOTA cpCER of 29.13% on this dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2401_03424
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MLCA-AVSR: Multi-Layer Cross Attention Fusion based Audio-Visual Speech Recognition
Wang, He
Guo, Pengcheng
Zhou, Pan
Xie, Lei
Sound
Artificial Intelligence
Audio and Speech Processing
While automatic speech recognition (ASR) systems degrade significantly in noisy environments, audio-visual speech recognition (AVSR) systems aim to complement the audio stream with noise-invariant visual cues and improve the system's robustness. However, current studies mainly focus on fusing the well-learned modality features, like the output of modality-specific encoders, without considering the contextual relationship during the modality feature learning. In this study, we propose a multi-layer cross-attention fusion based AVSR (MLCA-AVSR) approach that promotes representation learning of each modality by fusing them at different levels of audio/visual encoders. Experimental results on the MISP2022-AVSR Challenge dataset show the efficacy of our proposed system, achieving a concatenated minimum permutation character error rate (cpCER) of 30.57% on the Eval set and yielding up to 3.17% relative improvement compared with our previous system which ranked the second place in the challenge. Following the fusion of multiple systems, our proposed approach surpasses the first-place system, establishing a new SOTA cpCER of 29.13% on this dataset.
title MLCA-AVSR: Multi-Layer Cross Attention Fusion based Audio-Visual Speech Recognition
topic Sound
Artificial Intelligence
Audio and Speech Processing
url https://arxiv.org/abs/2401.03424