---
_id: '11809'
abstract:
- lang: eng
  text: This paper presents an end-to-end training approach for a beamformer-supported
    multi-channel ASR system. A neural network which estimates masks for a statistically
    optimum beamformer is jointly trained with a network for acoustic modeling. To
    update its parameters, we propagate the gradients from the acoustic model all
    the way through feature extraction and the complex valued beamforming operation.
    Besides avoiding a mismatch between the front-end and the back-end, this approach
    also eliminates the need for stereo data, i.e., the parallel availability of clean
    and noisy versions of the signals. Instead, it can be trained with real noisy
    multichannel data only. Also, relying on the signal statistics for beamforming,
    the approach makes no assumptions on the configuration of the microphone array.
    We further observe a performance gain through joint training in terms of word
    error rate in an evaluation of the system on the CHiME 4 dataset.
author:
- first_name: Jahn
  full_name: Heymann, Jahn
  id: '9168'
  last_name: Heymann
- first_name: Lukas
  full_name: Drude, Lukas
  id: '11213'
  last_name: Drude
- first_name: Christoph
  full_name: Boeddeker, Christoph
  id: '40767'
  last_name: Boeddeker
- first_name: Patrick
  full_name: Hanebrink, Patrick
  last_name: Hanebrink
- first_name: Reinhold
  full_name: Haeb-Umbach, Reinhold
  id: '242'
  last_name: Haeb-Umbach
citation:
  ama: 'Heymann J, Drude L, Boeddeker C, Hanebrink P, Haeb-Umbach R. BEAMNET: End-to-End
    Training of a Beamformer-Supported Multi-Channel ASR System. In: <i>Proc. IEEE
    Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP)</i>. ; 2017.'
  apa: 'Heymann, J., Drude, L., Boeddeker, C., Hanebrink, P., &#38; Haeb-Umbach, R.
    (2017). BEAMNET: End-to-End Training of a Beamformer-Supported Multi-Channel ASR
    System. In <i>Proc. IEEE Intl. Conf. on Acoustics, Speech and Signal Processing
    (ICASSP)</i>.'
  bibtex: '@inproceedings{Heymann_Drude_Boeddeker_Hanebrink_Haeb-Umbach_2017, title={BEAMNET:
    End-to-End Training of a Beamformer-Supported Multi-Channel ASR System}, booktitle={Proc.
    IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP)}, author={Heymann,
    Jahn and Drude, Lukas and Boeddeker, Christoph and Hanebrink, Patrick and Haeb-Umbach,
    Reinhold}, year={2017} }'
  chicago: 'Heymann, Jahn, Lukas Drude, Christoph Boeddeker, Patrick Hanebrink, and
    Reinhold Haeb-Umbach. “BEAMNET: End-to-End Training of a Beamformer-Supported
    Multi-Channel ASR System.” In <i>Proc. IEEE Intl. Conf. on Acoustics, Speech and
    Signal Processing (ICASSP)</i>, 2017.'
  ieee: 'J. Heymann, L. Drude, C. Boeddeker, P. Hanebrink, and R. Haeb-Umbach, “BEAMNET:
    End-to-End Training of a Beamformer-Supported Multi-Channel ASR System,” in <i>Proc.
    IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP)</i>, 2017.'
  mla: 'Heymann, Jahn, et al. “BEAMNET: End-to-End Training of a Beamformer-Supported
    Multi-Channel ASR System.” <i>Proc. IEEE Intl. Conf. on Acoustics, Speech and
    Signal Processing (ICASSP)</i>, 2017.'
  short: 'J. Heymann, L. Drude, C. Boeddeker, P. Hanebrink, R. Haeb-Umbach, in: Proc.
    IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP), 2017.'
date_created: 2019-07-12T05:28:40Z
date_updated: 2022-01-06T06:51:09Z
department:
- _id: '54'
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://groups.uni-paderborn.de/nt/pubs/2017/icassp_2017_heymann_paper.pdf
oa: '1'
project:
- _id: '52'
  name: Computing Resources Provided by the Paderborn Center for Parallel Computing
publication: Proc. IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP)
related_material:
  link:
  - description: Poster
    relation: supplementary_material
    url: https://groups.uni-paderborn.de/nt/pubs/2017/icassp_2017_heymann_poster.pdf
status: public
title: 'BEAMNET: End-to-End Training of a Beamformer-Supported Multi-Channel ASR System'
type: conference
user_id: '40767'
year: '2017'
...