The RWTH/UPB system combination for the CHiME 2018 Workshop

Kitza, Markus; Michel, Wilfried; Boeddeker, Christoph; Heitkaemper, Jens; Menne, Tobias; Schlüter, Ralf; Ney, Hermann; Schmalenstroeer, Joerg; Drude, Lukas; Heymann, Jahn; Haeb‐Umbach, Reinhold

doi:10.21437/chime.2018-12

Cited by 8 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the time of the challenge, Hitachi provided many contributions with Johns Hopkins University (JHU) on acoustic modeling (AM), language modeling (LM), and decoding techniques and achieved the second best result of a 48.2% WER [10]. On the other hand, Paderborn University achieved very promising speech enhancement (SE) techniques, named guided source separation (GSS) 2 , which achieved a significant improvement for evaluation data in multiple array settings [16,17]. We thought this is worth investigating to evaluate the results combining our contributions to assess the state-of-the-art performance of today's ASR system.…”

Section: Introductionmentioning

confidence: 99%

Guided Source Separation Meets a Strong ASR Backend: Hitachi/Paderborn University Joint Investigation for Dinner Party ASR

Kanda¹,

Böddeker²,

Heitkaemper³

et al. 2019

Interspeech 2019

Self Cite

View full text Add to dashboard Cite

In this paper, we present Hitachi and Paderborn University's joint effort for automatic speech recognition (ASR) in a dinner party scenario. The main challenges of ASR systems for dinner party recordings obtained by multiple microphone arrays are (1) heavy speech overlaps, (2) severe noise and reverberation, (3) very natural conversational content, and possibly (4) insufficient training data. As an example of a dinner party scenario, we have chosen the data presented during the CHiME-5 speech recognition challenge, where the baseline ASR had a 73.3% word error rate (WER), and even the best performing system at the CHiME-5 challenge had a 46.1% WER. We extensively investigated a combination of the guided source separation-based speech enhancement technique and an already proposed strong ASR backend and found that a tight combination of these techniques provided substantial accuracy improvements. Our final system achieved WERs of 39.94% and 41.64% for the development and evaluation data, respectively, both of which are the best published results for the dataset. We also investigated with additional training data on the official small data in the CHiME-5 corpus to assess the intrinsic difficulty of this ASR task.

show abstract

Section: Introductionmentioning

confidence: 99%

Guided Source Separation Meets a Strong ASR Backend: Hitachi/Paderborn University Joint Investigation for Dinner Party ASR

Kanda¹,

Böddeker²,

Heitkaemper³

et al. 2019

Interspeech 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Those WER improvements have been obtained using the acoustic model provided by the challenge organizers. Further significant gains are obtained if a stronger back-end and if system combination is used, see our companion paper [17]. An implementation of the described speech enhancement system without SAD is available on GitHub 4 .…”

Section: Discussionmentioning

confidence: 99%

“…We evaluate the proposed front-end processing techniques by computing Word Error Rates (WERs) using the baseline ASR back-end provided by the challenge organizers, and observe significant WER improvements. Yet better WERs can be achieved by combining the presented front-end with a stronger back-end, as is shown in [17].…”

Section: Introductionmentioning

confidence: 96%

Front-end processing for the CHiME-5 dinner party scenario

Boeddeker¹,

Heitkaemper²,

Schmalenstroeer³

et al. 2018

5th International Workshop on Speech Processing in Everyday Environments (CHiME 2018)

Self Cite

View full text Add to dashboard Cite

This contribution presents a speech enhancement system for the CHiME-5 Dinner Party Scenario. The front-end employs multi-channel linear time-variant filtering and achieves its gains without the use of a neural network. We present an adaptation of blind source separation techniques to the CHiME-5 database which we call Guided Source Separation (GSS). Using the baseline acoustic and language model, the combination of Weighted Prediction Error based dereverberation, guided source separation, and beamforming reduces the WER by 10.54 % (relative) for the single array track and by 21.12 % (relative) on the multiple array track.

show abstract

“…Although it's inferior to the USTC-iFlytek's, our system perform separation only once and has low computational complexity and model size apparently. The details on each ses- [32] 62.09 Toshiba [33] 63.30 STC [23] 63.30 RWTH-Paderborn [34] 68.40 Official [13] 80.28 sion and location over official AM and ours are given in Table 6. Even based on the official backend, our SD separation frontend contributes a 10% WER reduction, which is a significant improvement on this challenging task.…”

Section: Speaker-aware Trainingmentioning

confidence: 99%

Improved Speaker-Dependent Separation for CHiME-5 Challenge

Zhang

et al. 2019

Interspeech 2019

View full text Add to dashboard Cite

This paper summarizes several follow-up contributions for improving our submitted NWPU speaker-dependent system for CHiME-5 challenge, which aims to solve the problem of multi-channel, highly-overlapped conversational speech recognition in a dinner party scenario with reverberations and nonstationary noises. We adopt a speaker-aware training method by using i-vector as the target speaker information for multi-talker speech separation. With only one unified separation model for all speakers, we achieve a 10% absolute improvement in terms of word error rate (WER) over the previous baseline of 80.28% on the development set by leveraging our newly proposed data processing techniques and beamforming approach. With our improved back-end acoustic model, we further reduce WER to 60.15% which surpasses the result of our submitted CHiME-5 challenge system without applying any fusion techniques. Index Terms: CHiME-5 challenge, speaker-dependent speech separation, robust speech recognition, speech enhancement, beamforming * This work was done when the first author was an intern in Tencent AI lab.

show abstract

The RWTH/UPB system combination for the CHiME 2018 Workshop

Cited by 8 publications

References 14 publications

Guided Source Separation Meets a Strong ASR Backend: Hitachi/Paderborn University Joint Investigation for Dinner Party ASR

Guided Source Separation Meets a Strong ASR Backend: Hitachi/Paderborn University Joint Investigation for Dinner Party ASR

Front-end processing for the CHiME-5 dinner party scenario

Improved Speaker-Dependent Separation for CHiME-5 Challenge

Contact Info

Product

Resources

About