# Combining Frame-Synchronous and Label-Synchronous Systems for Speech Recognition

arXiv: https://arxiv.org/abs/2107.00764 · HTML: https://mlsanity.com/paper/2107.00764/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Jul 2021.

## Built on by (most cited first, 3 of 3)

- [Adapting GPT, GPT-2 and BERT Language Models for Speech Recognition](https://mlsanity.com/paper/2108.07789/) (2021)
- [Have best of both worlds: two-pass hybrid and E2E cascading framework for speech recognition](https://mlsanity.com/paper/2110.04891/) (2021)
- [Speaker- and Text-Independent Estimation of Articulatory Movements and Phoneme Alignments from Speech](https://mlsanity.com/paper/2407.03132/) (2024)

## Abstract (arXiv)

Commonly used automatic speech recognition (ASR) systems can be classified into frame-synchronous and label-synchronous categories, based on whether the speech is decoded on a per-frame or per-label basis. Frame-synchronous systems, such as traditional hidden Markov model systems, can easily incorporate existing knowledge and can support streaming ASR applications. Label-synchronous systems, based on attention-based encoder-decoder models, can jointly learn the acoustic and language information with a single model, which can be regarded as audio-grounded language models. In this paper, we propose rescoring the N-best hypotheses or lattices produced by a first-pass frame-synchronous system with a label-synchronous system in a second-pass. By exploiting the complementary modelling of the different approaches, the combined two-pass systems achieve competitive performance without using any extra speech or text data on two standard ASR tasks. For the 80-hour AMI IHM dataset, the combined system has a 13.7% word error rate (WER) on the evaluation set, which is up to a 29% relative WER reduction over the individual systems. For the 300-hour Switchboard dataset, the WERs of the combined system are 5.7% and 12.1% on Switchboard and CallHome subsets of Hub5'00, and 13.2% and 7.6% on Switchboard Cellular and Fisher subsets of RT03, up to a 33% relative reduction in WER over the individual systems.