# Multi-head Monotonic Chunkwise Attention For Online Speech Recognition

arXiv: https://arxiv.org/abs/2005.00205 · HTML: https://mlsanity.com/paper/2005.00205/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 May 2020.

## Built on by (most cited first, 5 of 5)

- [A Vector Quantized Approach for Text to Speech Synthesis on Real-World Spontaneous Speech](https://mlsanity.com/paper/2302.04215/) (2023)
- [Consistent Training and Decoding For End-to-end Speech Recognition Using Lattice-free MMI](https://mlsanity.com/paper/2112.02498/) (2021)
- [Decoupling recognition and transcription in Mandarin ASR](https://mlsanity.com/paper/2108.01129/) (2021)
- [Thank you for Attention: A survey on Attention-based Artificial Neural Networks for Automatic Speech Recognition](https://mlsanity.com/paper/2102.07259/) (2021)
- [On Biasing Transformer Attention Towards Monotonicity](https://mlsanity.com/paper/2104.03945/) (2021)

## Abstract (arXiv)

The attention mechanism of the Listen, Attend and Spell (LAS) model requires the whole input sequence to calculate the attention context and thus is not suitable for online speech recognition. To deal with this problem, we propose multi-head monotonic chunk-wise attention (MTH-MoChA), an improved version of MoChA. MTH-MoChA splits the input sequence into small chunks and computes multi-head attentions over the chunks. We also explore useful training strategies such as LSTM pooling, minimum world error rate training and SpecAugment to further improve the performance of MTH-MoChA. Experiments on AISHELL-1 data show that the proposed model, along with the training strategies, improve the character error rate (CER) of MoChA from 8.96% to 7.68% on test set. On another 18000 hours in-car speech data set, MTH-MoChA obtains 7.28% CER, which is significantly better than a state-of-the-art hybrid system.