pith. sign in

arxiv: 1612.02695 · v1 · pith:YSKZ2DKUnew · submitted 2016-12-08 · 💻 cs.NE · cs.CL· cs.LG· stat.ML

Towards better decoding and language model integration in sequence to sequence models

Jan Chorowski , Navdeep Jaitly This is my paper
classification 💻 cs.NE cs.CLcs.LGstat.ML
keywords languagemodelsmodelreachrecognitionseq2seqsequencespeech
0
0 comments X
read the original abstract

The recently proposed Sequence-to-Sequence (seq2seq) framework advocates replacing complex data processing pipelines, such as an entire automatic speech recognition system, with a single neural network trained in an end-to-end fashion. In this contribution, we analyse an attention-based seq2seq speech recognition system that directly transcribes recordings into characters. We observe two shortcomings: overconfidence in its predictions and a tendency to produce incomplete transcriptions when language models are used. We propose practical solutions to both problems achieving competitive speaker independent word error rates on the Wall Street Journal dataset: without separate language models we reach 10.6% WER, while together with a trigram language model, we reach 6.7% WER.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Response Time Enhances Alignment with Heterogeneous Preferences

    cs.LG 2026-05 unverdicted novelty 6.0

    Response times modeled as drift-diffusion processes enable consistent estimation of population-average preferences from heterogeneous anonymous binary choices.

  2. Non-Intrusive Automatic Speech Recognition Refinement: A Survey

    eess.AS 2025-08 accept novelty 4.0

    A survey that classifies non-intrusive ASR refinement methods into five categories, reviews domain adaptation and evaluation datasets, proposes standardized metrics, and identifies future research directions.

  3. Listen, Attend, Spell and Adapt: Speaker Adapted Sequence-to-Sequence ASR

    eess.AS 2019-07 unverdicted novelty 4.0

    KLD-based speaker adaptation of seq2seq ASR achieves 25% relative WER reduction, outperforming the 18.7% gain from conventional acoustic model adaptation.

  4. Integration of TensorFlow based Acoustic Model with Kaldi WFST Decoder

    eess.AS 2019-06 unverdicted novelty 3.0

    TensorFlow acoustic models are integrated with Kaldi WFST decoder and achieve equivalent word error rates to native Kaldi models on RM, WSJ, and LibriSpeech.