Pith. sign in

REVIEW 4 cited by

Transformer-Transducer: End-to-End Speech Recognition with Self-Attention

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

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 1910.12977 v1 pith:GKUWYLLA submitted 2019-10-28 eess.AS cs.CLcs.SD

Ching-Feng Yeh , Jay Mahadeokar , Kaustubh Kalgaonkar , Yongqiang Wang , Duc Le , Mahaveer Jain , Kjell Schubert , Christian Fuegen
show 1 more author
Michael L. Seltzer
This is my paper
classification eess.AScs.CLcs.SD
keywords networksself-attentiontransformercomplexityefficientend-to-endneuralrecognition
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We explore options to use Transformer networks in neural transducer for end-to-end speech recognition. Transformer networks use self-attention for sequence modeling and comes with advantages in parallel computation and capturing contexts. We propose 1) using VGGNet with causal convolution to incorporate positional information and reduce frame rate for efficient inference 2) using truncated self-attention to enable streaming for Transformer and reduce computational complexity. All experiments are conducted on the public LibriSpeech corpus. The proposed Transformer-Transducer outperforms neural transducer with LSTM/BLSTM networks and achieved word error rates of 6.37 % on the test-clean set and 15.30 % on the test-other set, while remaining streamable, compact with 45.7M parameters for the entire system, and computationally efficient with complexity of O(T), where T is input sequence length.

Discussion (0). Sign in to comment.

Forward citations

Cited by 4 Pith papers

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

  1. Progressive Alignment Objectives for Aligner-Encoder based ASR

    eess.AS 2026-06 unverdicted novelty 6.0 of 10

    InterAligner and InterCTC enable progressive alignment in Aligner-Encoder ASR, lowering LibriSpeech WER from 5.0/7.8 to 3.1/5.6 with gains on long utterances.

  2. WhisperRT -- Turning Whisper into a Causal Streaming Model

    cs.CL 2025-08 conditional novelty 6.0 of 10

    WhisperRT converts Whisper to a causal streaming ASR model via encoder causality, decoder synchronization on partial states, and fine-tuning, achieving better performance than non-fine-tuned streaming methods on sub-3...

  3. Progressive Alignment Objectives for Aligner-Encoder based ASR

    eess.AS 2026-06 unverdicted novelty 5.0 of 10

    InterAligner adds progressive intermediate alignment objectives plus InterCTC to Aligner-Encoder ASR, cutting WER from 5.0/7.8 to 3.1/5.6 on LibriSpeech test-clean/other with largest gains on long utterances.

  4. Exploring Cross-Utterance Speech Contexts for Conformer-Transducer Speech Recognition Systems

    eess.AS 2025-08 conditional novelty 5.0 of 10

    Adding cross-utterance audio context to Conformer-Transducer ASR reduces WER/CER by 0.5 to 1.1 absolute points on four benchmarks, and a splicing-based batch scheme cuts training time by up to about 19%.

Pith tools