Pith. sign in

REVIEW 1 cited by

Multi-blank Transducers for Speech Recognition

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2211.03541 v2 pith:5QZIMBMA submitted 2022-11-04 eess.AS cs.LGcs.SD

classification eess.AScs.LGcs.SD
keywords methodmulti-blankrnn-tblankblanksdatasetsinputlibrispeech
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

This paper proposes a modification to RNN-Transducer (RNN-T) models for automatic speech recognition (ASR). In standard RNN-T, the emission of a blank symbol consumes exactly one input frame; in our proposed method, we introduce additional blank symbols, which consume two or more input frames when emitted. We refer to the added symbols as big blanks, and the method multi-blank RNN-T. For training multi-blank RNN-Ts, we propose a novel logit under-normalization method in order to prioritize emissions of big blanks. With experiments on multiple languages and datasets, we show that multi-blank RNN-T methods could bring relative speedups of over +90%/+139% to model inference for English Librispeech and German Multilingual Librispeech datasets, respectively. The multi-blank RNN-T method also improves ASR accuracy consistently. We will release our implementation of the method in the NeMo (https://github.com/NVIDIA/NeMo) toolkit.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. WIND: Accelerated RNN-T Decoding with Windowed Inference for Non-blank Detection

    cs.LG 2025-05 conditional novelty 6.0 of 10

    WIND runs RNN-T decoding faster by evaluating windows of frames in parallel to find the first non-blank prediction, exactly matching greedy accuracy while running up to 2.4X faster.

Pith tools