Hands-free human-robot communication robust to speaker's radial position

Randy Gomez, Keisuke Nakamura, Kazuhiro Nakadai, Ui Hyun Kim, Hiroshi G. Okuno, Tatsuya Kawahara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

In this paper we present a method in room transfer function (RTF) estimation, employed specifically for dereverberation in hands-free human-robot communication.We introduce a radial distance compensation scheme which significantly improved the RTF estimate robust to the speech power variation due to changes in speaker's radial position. The proposed method is implemented in two levels; first, waveform-level compensation is executed to reflect the change in power caused by the change of radial position to the RTF. We generated possible RTF estimates within a close neighbourhood based on curve fitting. Then, we select among these estimates the optimal RTF based on acoustic model likelihood criterion, the same criterion employed in automatic speech recognition (ASR) systems. The latter is referred to as acoustic model-level compensation, which links the generated RTF to the ASR. We note that in ASR application, both waveform and acoustic models play an important role in achieving optimal performance. Thus, the synergistic effect of the two processes guarantee ASR performance improvement when used in conjunction with our ASR-based dereverberation scheme. Experimental evaluation show robustness in recognition performance when used in hands-free human-robot communication environment.

Original languageEnglish
Title of host publicationProceedings - IEEE International Conference on Robotics and Automation
Pages4329-4334
Number of pages6
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event2013 IEEE International Conference on Robotics and Automation, ICRA 2013 - Karlsruhe
Duration: 2013 May 62013 May 10

Other

Other2013 IEEE International Conference on Robotics and Automation, ICRA 2013
CityKarlsruhe
Period13/5/613/5/10

Fingerprint

End effectors
Transfer functions
Speech recognition
Robots
Communication
Acoustics
Curve fitting
Compensation and Redress

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Gomez, R., Nakamura, K., Nakadai, K., Kim, U. H., Okuno, H. G., & Kawahara, T. (2013). Hands-free human-robot communication robust to speaker's radial position. In Proceedings - IEEE International Conference on Robotics and Automation (pp. 4329-4334). [6631190] https://doi.org/10.1109/ICRA.2013.6631190

Hands-free human-robot communication robust to speaker's radial position. / Gomez, Randy; Nakamura, Keisuke; Nakadai, Kazuhiro; Kim, Ui Hyun; Okuno, Hiroshi G.; Kawahara, Tatsuya.

Proceedings - IEEE International Conference on Robotics and Automation. 2013. p. 4329-4334 6631190.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gomez, R, Nakamura, K, Nakadai, K, Kim, UH, Okuno, HG & Kawahara, T 2013, Hands-free human-robot communication robust to speaker's radial position. in Proceedings - IEEE International Conference on Robotics and Automation., 6631190, pp. 4329-4334, 2013 IEEE International Conference on Robotics and Automation, ICRA 2013, Karlsruhe, 13/5/6. https://doi.org/10.1109/ICRA.2013.6631190
Gomez R, Nakamura K, Nakadai K, Kim UH, Okuno HG, Kawahara T. Hands-free human-robot communication robust to speaker's radial position. In Proceedings - IEEE International Conference on Robotics and Automation. 2013. p. 4329-4334. 6631190 https://doi.org/10.1109/ICRA.2013.6631190
Gomez, Randy ; Nakamura, Keisuke ; Nakadai, Kazuhiro ; Kim, Ui Hyun ; Okuno, Hiroshi G. ; Kawahara, Tatsuya. / Hands-free human-robot communication robust to speaker's radial position. Proceedings - IEEE International Conference on Robotics and Automation. 2013. pp. 4329-4334
@inproceedings{197821ddb2c14f55a7bca9829c07384f,
title = "Hands-free human-robot communication robust to speaker's radial position",
abstract = "In this paper we present a method in room transfer function (RTF) estimation, employed specifically for dereverberation in hands-free human-robot communication.We introduce a radial distance compensation scheme which significantly improved the RTF estimate robust to the speech power variation due to changes in speaker's radial position. The proposed method is implemented in two levels; first, waveform-level compensation is executed to reflect the change in power caused by the change of radial position to the RTF. We generated possible RTF estimates within a close neighbourhood based on curve fitting. Then, we select among these estimates the optimal RTF based on acoustic model likelihood criterion, the same criterion employed in automatic speech recognition (ASR) systems. The latter is referred to as acoustic model-level compensation, which links the generated RTF to the ASR. We note that in ASR application, both waveform and acoustic models play an important role in achieving optimal performance. Thus, the synergistic effect of the two processes guarantee ASR performance improvement when used in conjunction with our ASR-based dereverberation scheme. Experimental evaluation show robustness in recognition performance when used in hands-free human-robot communication environment.",
author = "Randy Gomez and Keisuke Nakamura and Kazuhiro Nakadai and Kim, {Ui Hyun} and Okuno, {Hiroshi G.} and Tatsuya Kawahara",
year = "2013",
doi = "10.1109/ICRA.2013.6631190",
language = "English",
isbn = "9781467356411",
pages = "4329--4334",
booktitle = "Proceedings - IEEE International Conference on Robotics and Automation",

}

TY - GEN

T1 - Hands-free human-robot communication robust to speaker's radial position

AU - Gomez, Randy

AU - Nakamura, Keisuke

AU - Nakadai, Kazuhiro

AU - Kim, Ui Hyun

AU - Okuno, Hiroshi G.

AU - Kawahara, Tatsuya

PY - 2013

Y1 - 2013

N2 - In this paper we present a method in room transfer function (RTF) estimation, employed specifically for dereverberation in hands-free human-robot communication.We introduce a radial distance compensation scheme which significantly improved the RTF estimate robust to the speech power variation due to changes in speaker's radial position. The proposed method is implemented in two levels; first, waveform-level compensation is executed to reflect the change in power caused by the change of radial position to the RTF. We generated possible RTF estimates within a close neighbourhood based on curve fitting. Then, we select among these estimates the optimal RTF based on acoustic model likelihood criterion, the same criterion employed in automatic speech recognition (ASR) systems. The latter is referred to as acoustic model-level compensation, which links the generated RTF to the ASR. We note that in ASR application, both waveform and acoustic models play an important role in achieving optimal performance. Thus, the synergistic effect of the two processes guarantee ASR performance improvement when used in conjunction with our ASR-based dereverberation scheme. Experimental evaluation show robustness in recognition performance when used in hands-free human-robot communication environment.

AB - In this paper we present a method in room transfer function (RTF) estimation, employed specifically for dereverberation in hands-free human-robot communication.We introduce a radial distance compensation scheme which significantly improved the RTF estimate robust to the speech power variation due to changes in speaker's radial position. The proposed method is implemented in two levels; first, waveform-level compensation is executed to reflect the change in power caused by the change of radial position to the RTF. We generated possible RTF estimates within a close neighbourhood based on curve fitting. Then, we select among these estimates the optimal RTF based on acoustic model likelihood criterion, the same criterion employed in automatic speech recognition (ASR) systems. The latter is referred to as acoustic model-level compensation, which links the generated RTF to the ASR. We note that in ASR application, both waveform and acoustic models play an important role in achieving optimal performance. Thus, the synergistic effect of the two processes guarantee ASR performance improvement when used in conjunction with our ASR-based dereverberation scheme. Experimental evaluation show robustness in recognition performance when used in hands-free human-robot communication environment.

UR - http://www.scopus.com/inward/record.url?scp=84887303751&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887303751&partnerID=8YFLogxK

U2 - 10.1109/ICRA.2013.6631190

DO - 10.1109/ICRA.2013.6631190

M3 - Conference contribution

SN - 9781467356411

SP - 4329

EP - 4334

BT - Proceedings - IEEE International Conference on Robotics and Automation

ER -