# Teach an all-rounder with experts in different domains

## TL;DR

This paper introduces a multi-domain acoustic model trained via teacher-student framework, improving speech recognition accuracy across various speaking styles and noise conditions.

## Contribution

It proposes a novel multi-domain training approach using domain-specific teacher models to enhance a single student model's performance in diverse ASR tasks.

## Key findings

- Up to 10.4% relative CER improvement over baseline
- Effective across different speaking styles and noise conditions
- Applicable with DFSMN and LSTM models

## Abstract

In many automatic speech recognition (ASR) tasks, an ideal model has to be applicable over multiple domains. In this paper, we propose to teach an all-rounder with experts in different domains. Concretely, we build a multi-domain acoustic model by applying the teacher-student training framework. First, for each domain, a teacher model (domain-dependent model) is trained by fine-tuning a multi-condition model with domain-specific subset. Then all these teacher models are used to teach one single student model simultaneously. We perform experiments on two predefined domain setups. One is domains with different speaking styles, the other is nearfield, far-field and far-field with noise. Moreover, two types of models are examined: deep feedforward sequential memory network (DFSMN) and long short term memory (LSTM). Experimental results show that the model trained with this framework outperforms not only multi-condition model but also domain-dependent model. Specially, our training method provides up to 10.4% relative character error rate improvement over baseline model (multi-condition model).