Dyn-ASR: Compact, Multilingual Speech Recognition via Spoken Language and Accent Identification

2021/08/04 by Sangeeta Ghangam, Ghangam, Sangeeta, Daniel Whitenack +3
Computer Science · Engineering · #Audio and Speech Processing (eess.AS) #Computation and Language (cs.CL) #FOS: Computer and information sciences #FOS: Electrical engineering #Sound (cs.SD) #cs.CL #cs.SD #eess.AS #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2108.02034

Accepted to IEEE WF-IOT 2021

arxiv created 2021/08/04 · arxiv updated 2021/08/05

Abstract

Running automatic speech recognition (ASR) on edge devices is non-trivial due to resource constraints, especially in scenarios that require supporting multiple languages. We propose a new approach to enable multilingual speech recognition on edge devices. This approach uses both language identification and accent identification to select one of multiple monolingual ASR models on-the-fly, each fine-tuned for a particular accent. Initial results for both recognition performance and resource usage are promising with our approach using less than 1/12th of the memory consumed by other solutions.

Related

• Selective Invocation for Multilingual ASR: A Cost-effective Approach Adapting to Speech Recognition Difficulty
• Streaming Language Identification using Combination of Acoustic Representations and ASR Hypotheses
• Flavors of Moonshine: Tiny Specialized ASR Models for Edge Devices
• Multilingual Approach to Joint Speech and Accent Recognition with DNN-HMM Framework
• A Language Agnostic Multilingual Streaming On-Device ASR System