Directional Source Separation for Robust Speech Recognition on Smart Glasses

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Feng, Tiantian, Lin, Ju, Huang, Yiteng, He, Weipeng, Kalgaonkar, Kaustubh, Moritz, Niko, Wan, Li, Lei, Xin, Sun, Ming, Seide, Frank
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909648712368128
author Feng, Tiantian
Lin, Ju
Huang, Yiteng
He, Weipeng
Kalgaonkar, Kaustubh
Moritz, Niko
Wan, Li
Lei, Xin
Sun, Ming
Seide, Frank
author_facet Feng, Tiantian
Lin, Ju
Huang, Yiteng
He, Weipeng
Kalgaonkar, Kaustubh
Moritz, Niko
Wan, Li
Lei, Xin
Sun, Ming
Seide, Frank
contents Modern smart glasses leverage advanced audio sensing and machine learning technologies to offer real-time transcribing and captioning services, considerably enriching human experiences in daily communications. However, such systems frequently encounter challenges related to environmental noises, resulting in degradation to speech recognition and speaker change detection. To improve voice quality, this work investigates directional source separation using the multi-microphone array. We first explore multiple beamformers to assist source separation modeling by strengthening the directional properties of speech signals. In addition to relying on predetermined beamformers, we investigate neural beamforming in multi-channel source separation, demonstrating that automatic learning directional characteristics effectively improves separation quality. We further compare the ASR performance leveraging separated outputs to noisy inputs. Our results show that directional source separation benefits ASR for the wearer but not for the conversation partner. Lastly, we perform the joint training of the directional source separation and ASR model, achieving the best overall ASR performance.
format Preprint
id arxiv_https___arxiv_org_abs_2309_10993
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Directional Source Separation for Robust Speech Recognition on Smart Glasses
Feng, Tiantian
Lin, Ju
Huang, Yiteng
He, Weipeng
Kalgaonkar, Kaustubh
Moritz, Niko
Wan, Li
Lei, Xin
Sun, Ming
Seide, Frank
Sound
Human-Computer Interaction
Audio and Speech Processing
Modern smart glasses leverage advanced audio sensing and machine learning technologies to offer real-time transcribing and captioning services, considerably enriching human experiences in daily communications. However, such systems frequently encounter challenges related to environmental noises, resulting in degradation to speech recognition and speaker change detection. To improve voice quality, this work investigates directional source separation using the multi-microphone array. We first explore multiple beamformers to assist source separation modeling by strengthening the directional properties of speech signals. In addition to relying on predetermined beamformers, we investigate neural beamforming in multi-channel source separation, demonstrating that automatic learning directional characteristics effectively improves separation quality. We further compare the ASR performance leveraging separated outputs to noisy inputs. Our results show that directional source separation benefits ASR for the wearer but not for the conversation partner. Lastly, we perform the joint training of the directional source separation and ASR model, achieving the best overall ASR performance.
title Directional Source Separation for Robust Speech Recognition on Smart Glasses
topic Sound
Human-Computer Interaction
Audio and Speech Processing
url https://arxiv.org/abs/2309.10993