Vectorized beam search for CTC-attention-based speech recognition

Hiroshi Seki, Takaaki Hori, Shinji Watanabe, Niko Moritz, Jonathan Le Roux

Research output: Contribution to journalConference article

2 Citations (Scopus)

Abstract

This paper investigates efficient beam search techniques for end-to-end automatic speech recognition (ASR) with attention-based encoder-decoder architecture. We accelerate the decoding process by vectorizing multiple hypotheses during the beam search, where we replace the score accumulation steps for each hypothesis with vector-matrix operations for the vectorized hypotheses. This modification allows us to take advantage of the parallel computing capabilities of multi-core CPUs and GPUs, resulting in significant speedups and also enabling us to process multiple utterances in a batch simultaneously. Moreover, we extend the decoding method to incorporate a recurrent neural network language model (RNNLM) and connectionist temporal classification (CTC) scores, which typically improve ASR accuracy but have not been investigated for the use of such parallelized decoding algorithms. Experiments with LibriSpeech and Corpus of Spontaneous Japanese datasets have demonstrated that the vectorized beam search achieves 1.8× speedup on a CPU and 33× speedup on a GPU compared with the original CPU implementation. When using joint CTC/attention decoding with an RNNLM, we also achieved 11× speedup on a GPU while maintaining the benefits of CTC and RNNLM. With these benefits, we achieved almost real-time processing with a small latency of 0.1× real-time without streaming search process.

Original languageEnglish
Pages (from-to)3825-3829
Number of pages5
JournalProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Volume2019-September
DOIs
Publication statusPublished - 2019
Externally publishedYes
Event20th Annual Conference of the International Speech Communication Association: Crossroads of Speech and Language, INTERSPEECH 2019 - Graz, Austria
Duration: 2019 Sep 152019 Sep 19

Keywords

  • Beam search
  • Encoder-decoder network
  • GPU
  • Parallel computing
  • Speech recognition

ASJC Scopus subject areas

  • Language and Linguistics
  • Human-Computer Interaction
  • Signal Processing
  • Software
  • Modelling and Simulation

Fingerprint Dive into the research topics of 'Vectorized beam search for CTC-attention-based speech recognition'. Together they form a unique fingerprint.

  • Cite this