# Improving accuracy of rare words for RNN-Transducer through unigram shallow fusion

arXiv: https://arxiv.org/abs/2012.00133 · HTML: https://mlsanity.com/paper/2012.00133/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Nov 2020.

## Built on by (most cited first, 6 of 6)

- [Large-scale Language Model Rescoring on Long-form Data](https://mlsanity.com/paper/2306.08133/) (2023)
- [Fast and accurate factorized neural transducer for text adaption of end-to-end speech recognition models](https://mlsanity.com/paper/2212.01992/) (2022)
- [Adaptive Discounting of Implicit Language Models in RNN-Transducers](https://mlsanity.com/paper/2203.02317/) (2022)
- [Improving RNN-Transducers with Acoustic LookAhead](https://mlsanity.com/paper/2307.05006/) (2023)
- [Improving Rare Words Recognition through Homophone Extension and Unified Writing for Low-resource Cantonese Speech Recognition](https://mlsanity.com/paper/2302.00836/) (2023)
- [From Oracle to Noisy Context: Mitigating Contextual Exposure Bias in Speech-LLMs](https://mlsanity.com/paper/2603.24034/) (2026)

## Abstract (arXiv)

End-to-end automatic speech recognition (ASR) systems, such as recurrent neural network transducer (RNN-T), have become popular, but rare word remains a challenge. In this paper, we propose a simple, yet effective method called unigram shallow fusion (USF) to improve rare words for RNN-T. In USF, we extract rare words from RNN-T training data based on unigram count, and apply a fixed reward when the word is encountered during decoding. We show that this simple method can improve performance on rare words by 3.7% WER relative without degradation on general test set, and the improvement from USF is additive to any additional language model based rescoring. Then, we show that the same USF does not work on conventional hybrid system. Finally, we reason that USF works by fixing errors in probability estimates of words due to Viterbi search used during decoding with subword-based RNN-T.