Robust ego noise suppression of a robot

Gökhan Ince; Kazuhiro Nakadai; Tobias Rodemann; Hiroshi Tsujino; Jun Ichi Imura

doi:10.1007/978-3-642-13022-9_7

Robust ego noise suppression of a robot

Gökhan Ince^*, Kazuhiro Nakadai, Tobias Rodemann, Hiroshi Tsujino, Jun Ichi Imura

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Citations (Scopus)

Abstract

This paper describes an architecture that can enhance a robot with the capability of performing automatic speech recognition even while the robot is moving. The system consists of three blocks: (1) a multi-channel noise reduction block comprising consequent stages of microphone-array-based sound localization, geometric source separation and post filtering, (2) a single-channel template subtraction block and (3) a speech recognition block. In this work, we specifically investigate a missing feature theory based automatic speech recognition (MFT-ASR) approach in block (3), that makes use of spectrotemporal elements that are derived from (1) and (2) to measure the reliability of the audio features and to generate masks that filter unreliable speech features. We evaluate the proposed technique on a robot using word error rates. Furthermore, we present a detailed analysis of recognition accuracy to determine optimal parameters. Proposed MFT-ASR implementation attains significantly higher recognition performance compared to the performances of both single and multi-channel noise reduction methods.

Original language	English
Title of host publication	Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings
Pages	62-71
Number of pages	10
Edition	PART 1
DOIs	https://doi.org/10.1007/978-3-642-13022-9_7
Publication status	Published - 2010
Externally published	Yes
Event	23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010 - Cordoba, Spain Duration: 1 Jun 2010 → 4 Jun 2010

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 1
Volume	6096 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010
Country/Territory	Spain
City	Cordoba
Period	1/06/10 → 4/06/10

Keywords

Ego noise
microphone array
missing feature theory
noise reduction
robot audition
speech recognition

Access to Document

10.1007/978-3-642-13022-9_7

Cite this

Ince, G., Nakadai, K., Rodemann, T., Tsujino, H., & Imura, J. I. (2010). Robust ego noise suppression of a robot. In Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings (PART 1 ed., pp. 62-71). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6096 LNAI, No. PART 1). https://doi.org/10.1007/978-3-642-13022-9_7

Ince, Gökhan ; Nakadai, Kazuhiro ; Rodemann, Tobias et al. / Robust ego noise suppression of a robot. Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings. PART 1. ed. 2010. pp. 62-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).

@inproceedings{41edcef278234ccc99a3c536518fbd38,

title = "Robust ego noise suppression of a robot",

abstract = "This paper describes an architecture that can enhance a robot with the capability of performing automatic speech recognition even while the robot is moving. The system consists of three blocks: (1) a multi-channel noise reduction block comprising consequent stages of microphone-array-based sound localization, geometric source separation and post filtering, (2) a single-channel template subtraction block and (3) a speech recognition block. In this work, we specifically investigate a missing feature theory based automatic speech recognition (MFT-ASR) approach in block (3), that makes use of spectrotemporal elements that are derived from (1) and (2) to measure the reliability of the audio features and to generate masks that filter unreliable speech features. We evaluate the proposed technique on a robot using word error rates. Furthermore, we present a detailed analysis of recognition accuracy to determine optimal parameters. Proposed MFT-ASR implementation attains significantly higher recognition performance compared to the performances of both single and multi-channel noise reduction methods.",

keywords = "Ego noise, microphone array, missing feature theory, noise reduction, robot audition, speech recognition",

author = "G{\"o}khan Ince and Kazuhiro Nakadai and Tobias Rodemann and Hiroshi Tsujino and Imura, {Jun Ichi}",

year = "2010",

doi = "10.1007/978-3-642-13022-9_7",

language = "English",

isbn = "3642130216",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 1",

pages = "62--71",

booktitle = "Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings",

edition = "PART 1",

note = "23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010 ; Conference date: 01-06-2010 Through 04-06-2010",

}

Ince, G, Nakadai, K, Rodemann, T, Tsujino, H & Imura, JI 2010, Robust ego noise suppression of a robot. in Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 6096 LNAI, pp. 62-71, 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010, Cordoba, Spain, 1/06/10. https://doi.org/10.1007/978-3-642-13022-9_7

Robust ego noise suppression of a robot. / Ince, Gökhan; Nakadai, Kazuhiro; Rodemann, Tobias et al.
Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings. PART 1. ed. 2010. p. 62-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6096 LNAI, No. PART 1).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Robust ego noise suppression of a robot

AU - Ince, Gökhan

AU - Nakadai, Kazuhiro

AU - Rodemann, Tobias

AU - Tsujino, Hiroshi

AU - Imura, Jun Ichi

PY - 2010

Y1 - 2010

N2 - This paper describes an architecture that can enhance a robot with the capability of performing automatic speech recognition even while the robot is moving. The system consists of three blocks: (1) a multi-channel noise reduction block comprising consequent stages of microphone-array-based sound localization, geometric source separation and post filtering, (2) a single-channel template subtraction block and (3) a speech recognition block. In this work, we specifically investigate a missing feature theory based automatic speech recognition (MFT-ASR) approach in block (3), that makes use of spectrotemporal elements that are derived from (1) and (2) to measure the reliability of the audio features and to generate masks that filter unreliable speech features. We evaluate the proposed technique on a robot using word error rates. Furthermore, we present a detailed analysis of recognition accuracy to determine optimal parameters. Proposed MFT-ASR implementation attains significantly higher recognition performance compared to the performances of both single and multi-channel noise reduction methods.

AB - This paper describes an architecture that can enhance a robot with the capability of performing automatic speech recognition even while the robot is moving. The system consists of three blocks: (1) a multi-channel noise reduction block comprising consequent stages of microphone-array-based sound localization, geometric source separation and post filtering, (2) a single-channel template subtraction block and (3) a speech recognition block. In this work, we specifically investigate a missing feature theory based automatic speech recognition (MFT-ASR) approach in block (3), that makes use of spectrotemporal elements that are derived from (1) and (2) to measure the reliability of the audio features and to generate masks that filter unreliable speech features. We evaluate the proposed technique on a robot using word error rates. Furthermore, we present a detailed analysis of recognition accuracy to determine optimal parameters. Proposed MFT-ASR implementation attains significantly higher recognition performance compared to the performances of both single and multi-channel noise reduction methods.

KW - Ego noise

KW - microphone array

KW - missing feature theory

KW - noise reduction

KW - robot audition

KW - speech recognition

UR - http://www.scopus.com/inward/record.url?scp=78651479949&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-13022-9_7

DO - 10.1007/978-3-642-13022-9_7

M3 - Conference contribution

AN - SCOPUS:78651479949

SN - 3642130216

SN - 9783642130212

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 62

EP - 71

BT - Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings

T2 - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligence Systems, IEA/AIE 2010

Y2 - 1 June 2010 through 4 June 2010

ER -

Ince G, Nakadai K, Rodemann T, Tsujino H, Imura JI. Robust ego noise suppression of a robot. In Trends in Applied Intelligent Systems - 23rd International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2010, Proceedings. PART 1 ed. 2010. p. 62-71. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1). doi: 10.1007/978-3-642-13022-9_7

Robust ego noise suppression of a robot

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this