Less is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging

By Rohit Prabhavalkar, Yanzhang He, David Rybach, Sean Campbell, Arun Narayanan, Trevor Strohman, Tara N. Sainath. 2020

37 citations recorded by the source corpus.

End-to-end models that condition the output sequence on all previously predicted labels have emerged as popular alternatives to conventional systems for automatic speech recognition (ASR). Since distinct label histories correspond to distinct models states, such models are decoded using an approximate beam-search which produces a tree of hypotheses.In this work, we study the influence of the amount of label context on the model’s accuracy, and its impact on the efficiency of the decoding process. We find that we can limit the context of the recurrent neural network transducer (RNN-T) during training to just four previous word-piece labels, without degrading word error rate (WER) relative to the full-context baseline. Limiting context also provides opportunities to improve decoding efficiency by removing redundant paths from the active beam, and instead retaining them in the final lattice. This path-merging scheme can also be applied when decoding the baseline full-context model through an approximation. Overall, we find that the proposed path-merging scheme is extremely effective, allowing us to improve oracle WERs by up to 36% over the baseline, while simultaneously reducing the number of model evaluations by up to 5.3% without any degradation in WER, or up to 15.7% when lattice rescoring is applied.

Subfields: Speech & Audio

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Speech & Audio

Citations in the corpus

Multitask Training with Text Data for End-to-End Speech Recognition (2020, background): “Attention-based end-to-end (E2E) speech recognition systems map audio features directly to text-level representations [1, 2, 3, 4, 5].”
Recent Advances in End-to-End Automatic Speech Recognition (2021, background): “efficient decoding strategy, which was explored in [214, 240].”
End-to-End Speech Recognition: A Survey (2023, background): “simple embedding lookup table, that attends to only a few previous tokens from the model [34], [77], [78], [79], [80].”
Fast and Parallel Decoding for Transducer (2022, background): “Another way to reduce the number of active hypotheses is proposed in [5] by using a prediction network 1 https://github.com/k2-fsa/icefall with limited label context to merge hypotheses with identical prediction histories.”