We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: Exploiting Single-Channel Speech For Multi-channel End-to-end Speech Recognition

Abstract: Recently, the end-to-end training approach for neural beamformer-supported multi-channel ASR has shown its effectiveness in multi-channel speech recognition. However, the integration of multiple modules makes it more difficult to perform end-to-end training, particularly given that the multi-channel speech corpus recorded in real environments with a sizeable data scale is relatively limited. This paper explores the usage of single-channel data to improve the multi-channel end-to-end speech recognition system. Specifically, we design three schemes to exploit the single-channel data, namely pre-training, data scheduling, and data simulation. Extensive experiments on CHiME4 and AISHELL-4 datasets demonstrate that all three methods improve the multi-channel end-to-end training stability and speech recognition performance, while the data scheduling approach keeps a much simpler pipeline (vs. pre-training) and less computation cost (vs. data simulation). Moreover, we give a thorough analysis of our systems, including how the performance is affected by the choice of front-end, the data augmentation, training strategy, and single-channel data size.
Comments: submitted to ASRU 2021
Subjects: Audio and Speech Processing (eess.AS)
Cite as: arXiv:2107.02670 [eess.AS]
  (or arXiv:2107.02670v1 [eess.AS] for this version)

Submission history

From: Keyu An [view email]
[v1] Tue, 6 Jul 2021 15:21:41 GMT (411kb,D)

Link back to: arXiv, form interface, contact.