Speechyolo: Detection and localization of speech objects

Yael Segal; Tzeviya Sylvia Fuchs; Joseph Keshet

doi:10.21437/interspeech.2019-1749

Speechyolo: Detection and localization of speech objects

Yael Segal, Tzeviya Sylvia Fuchs, Joseph Keshet

Department of Computer Science

Research output: Contribution to journal › Conference article › peer-review

8 Scopus citations

Abstract

In this paper, we propose to apply object detection methods from the vision domain on the speech recognition domain, by treating audio fragments as objects. More specifically, we present SpeechYOLO, which is inspired by the YOLO algorithm [1] for object detection in images. The goal of SpeechYOLO is to localize boundaries of utterances within the input signal, and to correctly classify them. Our system is composed of a convolutional neural network, with a simple least-mean-squares loss function. We evaluated the system on several keyword spotting tasks, that include corpora of read speech and spontaneous speech. Our system compares favorably with other algorithms trained for both localization and classification.

Original language	English
Pages (from-to)	4210-4214
Number of pages	5
Journal	Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Volume	2019-September
DOIs	https://doi.org/10.21437/interspeech.2019-1749
State	Published - 2019
Event	20th Annual Conference of the International Speech Communication Association: Crossroads of Speech and Language, INTERSPEECH 2019 - Graz, Austria Duration: 15 Sep 2019 → 19 Sep 2019

Bibliographical note

Publisher Copyright:
Copyright © 2019 ISCA

Keywords

Convolutional neural networks
Event detection
Keyword spotting
Speech processing

Access to Document

10.21437/interspeech.2019-1749

Cite this

@article{a16729db3a7f4007a583e5795d6f888e,

title = "Speechyolo: Detection and localization of speech objects",

abstract = "In this paper, we propose to apply object detection methods from the vision domain on the speech recognition domain, by treating audio fragments as objects. More specifically, we present SpeechYOLO, which is inspired by the YOLO algorithm [1] for object detection in images. The goal of SpeechYOLO is to localize boundaries of utterances within the input signal, and to correctly classify them. Our system is composed of a convolutional neural network, with a simple least-mean-squares loss function. We evaluated the system on several keyword spotting tasks, that include corpora of read speech and spontaneous speech. Our system compares favorably with other algorithms trained for both localization and classification.",

keywords = "Convolutional neural networks, Event detection, Keyword spotting, Speech processing",

author = "Yael Segal and Fuchs, \{Tzeviya Sylvia\} and Joseph Keshet",

note = "Publisher Copyright: Copyright {\textcopyright} 2019 ISCA; 20th Annual Conference of the International Speech Communication Association: Crossroads of Speech and Language, INTERSPEECH 2019 ; Conference date: 15-09-2019 Through 19-09-2019",

year = "2019",

doi = "10.21437/interspeech.2019-1749",

language = "אנגלית",

volume = "2019-September",

pages = "4210--4214",

journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",

issn = "2308-457X",

publisher = "International Speech Communication Association",

}

TY - JOUR

T1 - Speechyolo

T2 - 20th Annual Conference of the International Speech Communication Association: Crossroads of Speech and Language, INTERSPEECH 2019

AU - Segal, Yael

AU - Fuchs, Tzeviya Sylvia

AU - Keshet, Joseph

PY - 2019

Y1 - 2019

N2 - In this paper, we propose to apply object detection methods from the vision domain on the speech recognition domain, by treating audio fragments as objects. More specifically, we present SpeechYOLO, which is inspired by the YOLO algorithm [1] for object detection in images. The goal of SpeechYOLO is to localize boundaries of utterances within the input signal, and to correctly classify them. Our system is composed of a convolutional neural network, with a simple least-mean-squares loss function. We evaluated the system on several keyword spotting tasks, that include corpora of read speech and spontaneous speech. Our system compares favorably with other algorithms trained for both localization and classification.

AB - In this paper, we propose to apply object detection methods from the vision domain on the speech recognition domain, by treating audio fragments as objects. More specifically, we present SpeechYOLO, which is inspired by the YOLO algorithm [1] for object detection in images. The goal of SpeechYOLO is to localize boundaries of utterances within the input signal, and to correctly classify them. Our system is composed of a convolutional neural network, with a simple least-mean-squares loss function. We evaluated the system on several keyword spotting tasks, that include corpora of read speech and spontaneous speech. Our system compares favorably with other algorithms trained for both localization and classification.

KW - Convolutional neural networks

KW - Event detection

KW - Keyword spotting

KW - Speech processing

UR - http://www.scopus.com/inward/record.url?scp=85074717664&partnerID=8YFLogxK

U2 - 10.21437/interspeech.2019-1749

DO - 10.21437/interspeech.2019-1749

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.conferencearticle???

AN - SCOPUS:85074717664

SN - 2308-457X

VL - 2019-September

SP - 4210

EP - 4214

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

Y2 - 15 September 2019 through 19 September 2019

ER -

Speechyolo: Detection and localization of speech objects

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this