Pith. sign in

REVIEW 5 cited by

Low-Latency Sequence-to-Sequence Speech Recognition and Translation by Partial Hypothesis Selection

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 2005.11185 v2 pith:ST3W25ZC submitted 2020-05-22 cs.CL cs.SDeess.AS

Danni Liu , Gerasimos Spanakis , Jan Niehues This is my paper · ORCID
classification cs.CLcs.SDeess.AS
keywords latencyspeechtranslationofflineapplicablebleucomparedencoder-decoder
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Encoder-decoder models provide a generic architecture for sequence-to-sequence tasks such as speech recognition and translation. While offline systems are often evaluated on quality metrics like word error rates (WER) and BLEU, latency is also a crucial factor in many practical use-cases. We propose three latency reduction techniques for chunk-based incremental inference and evaluate their efficiency in terms of accuracy-latency trade-off. On the 300-hour How2 dataset, we reduce latency by 83% to 0.8 second by sacrificing 1% WER (6% rel.) compared to offline transcription. Although our experiments use the Transformer, the hypothesis selection strategies are applicable to other encoder-decoder models. To avoid expensive re-computation, we use a unidirectionally-attending encoder. After an adaptation procedure to partial sequences, the unidirectional model performs on-par with the original model. We further show that our approach is also applicable to low-latency speech translation. On How2 English-Portuguese speech translation, we reduce latency to 0.7 second (-84% rel.) while incurring a loss of 2.4 BLEU points (5% rel.) compared to the offline system.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

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

  1. SimulS2ST-Omni: Data-Efficient Streaming Speech-to-Speech Translation via Explicit Trajectory Supervision

    cs.SD 2026-07 conditional novelty 6.0 of 10

    A joint text-code trajectory supervision recipe lets a two-stream speech LM achieve competitive long-form streaming S2ST with ~2k hours of paired speech.

  2. A Self-Training Approach for Whisper to Enhance Long Dysarthric Speech Recognition

    cs.SD 2025-06 conditional novelty 6.0 of 10

    An iterative segmentation-based self-training method for Whisper improved long dysarthric speech recognition and achieved second place in both WER and SemScore at the SAP Challenge.

  3. MFLA: Monotonic Finite Look-ahead Attention for Streaming Speech Recognition

    cs.CL 2025-06 conditional novelty 6.0 of 10

    MFLA adds finite look-ahead attention plus a CIF-based token counter to Whisper, enabling streaming recognition with a wait-k latency-quality trade-off.

  4. WhisperKit: On-device Real-time ASR with Billion-Scale Transformers

    cs.SD 2025-07 conditional novelty 5.0 of 10

    WhisperKit's optimized on-device Whisper Large v3 Turbo streaming system reportedly achieves 0.46 s per-word latency and 2.2% WER, beating cloud baselines in its benchmark.

  5. NIM4-ASR: Towards Efficient, Robust, and Customizable Real-Time LLM-Based ASR

    eess.AS 2026-04 unverdicted novelty 4.0 of 10

    NIM4-ASR delivers SOTA ASR performance on public benchmarks using a 2.3B-parameter LLM with multi-stage training, real-time streaming, and million-scale hotword customization via RAG.

Pith tools