DeepAI AI Chat
Log In Sign Up

Unified Modeling of Multi-Domain Multi-Device ASR Systems

by   Soumyajit Mitra, et al.

Modern Automatic Speech Recognition (ASR) systems often use a portfolio of domain-specific models in order to get high accuracy for distinct user utterance types across different devices. In this paper, we propose an innovative approach that integrates the different per-domain per-device models into a unified model, using a combination of domain embedding, domain experts, mixture of experts and adversarial training. We run careful ablation studies to show the benefit of each of these innovations in contributing to the accuracy of the overall unified model. Experiments show that our proposed unified modeling approach actually outperforms the carefully tuned per-domain models, giving relative gains of up to 10 increase in the number of parameters.


page 1

page 2

page 3

page 4


A Deep Learning System for Domain-specific speech Recognition

As human-machine voice interfaces provide easy access to increasingly in...

On-Device Personalization of Automatic Speech Recognition Models for Disordered Speech

While current state-of-the-art Automatic Speech Recognition (ASR) system...

A practical framework for multi-domain speech recognition and an instance sampling method to neural language modeling

Automatic speech recognition (ASR) systems used on smart phones or vehic...

Efficient domain adaptation of language models in ASR systems using Prompt-tuning

Automatic Speech Recognition (ASR) systems have found their use in numer...

Domain Prompts: Towards memory and compute efficient domain adaptation of ASR systems

Automatic Speech Recognition (ASR) systems have found their use in numer...

Flexi-Transducer: Optimizing Latency, Accuracy and Compute forMulti-Domain On-Device Scenarios

Often, the storage and computational constraints of embeddeddevices dema...

Attention based end to end Speech Recognition for Voice Search in Hindi and English

We describe here our work with automatic speech recognition (ASR) in the...