English

NGPU-LM: GPU-Accelerated N-Gram Language Model for Context-Biasing in Greedy ASR Decoding

Audio and Speech Processing 2025-05-30 v1 Artificial Intelligence Computation and Language Machine Learning Sound

Abstract

Statistical n-gram language models are widely used for context-biasing tasks in Automatic Speech Recognition (ASR). However, existing implementations lack computational efficiency due to poor parallelization, making context-biasing less appealing for industrial use. This work rethinks data structures for statistical n-gram language models to enable fast and parallel operations for GPU-optimized inference. Our approach, named NGPU-LM, introduces customizable greedy decoding for all major ASR model types - including transducers, attention encoder-decoder models, and CTC - with less than 7% computational overhead. The proposed approach can eliminate more than 50% of the accuracy gap between greedy and beam search for out-of-domain scenarios while avoiding significant slowdown caused by beam search. The implementation of the proposed NGPU-LM is open-sourced.

Keywords

Cite

@article{arxiv.2505.22857,
  title  = {NGPU-LM: GPU-Accelerated N-Gram Language Model for Context-Biasing in Greedy ASR Decoding},
  author = {Vladimir Bataev and Andrei Andrusenko and Lilit Grigoryan and Aleksandr Laptev and Vitaly Lavrukhin and Boris Ginsburg},
  journal= {arXiv preprint arXiv:2505.22857},
  year   = {2025}
}

Comments

Accepted to Interspeech 2025

R2 v1 2026-07-01T02:47:22.701Z