Abstract: We investigate using large language models (LLMs) to generate text-only training data for improving multilingual and code-switching automatic speech ...
Code-switching text has been generated by first creating parallel text for two languages, followed by alignment and text re- placement, to improve ASR for text ...
We study the problem of multilingual masked language modeling, i.e. the training of a single model on concatenated text from multiple languages, and present a ...
We investigate methods to augment Code-switching training text data by artificially generating them. Concretely, we propose a cycle-consistent adversarial ...
Nov 6, 2024 · In this paper, we study the effectiveness of multilingual large language models (MLLMs) across monolingual (English-only, Spanish-only) and bilingual ( ...
Nov 4, 2022 · With the rise of globalisation, there is an increasing demand for multilingual ASR, handling language and dialectal variation of spoken ...
Oct 18, 2024 · This paper leverages advancements in large language models. (LLMs) to enhance CS-ASR systems by generating Arabic-. English code-switched ...
Sep 16, 2023 · This paper proposes (1) a new method for creating code-switching. ASR datasets from purely monolingual data sources, and (2) a novel ...
People also ask
Are multilingual models effective in code-switching?
What is language model in ASR?
What is code-switching in speech recognition?
What do you understand by code-switching?
Improving Multilingual and Code-Switching ASR Using Large Language Model Generated Text · Aligning Speech to Languages to Enhance Code-switching Speech ...
Sep 5, 2024 · In this work, a novel TTS model based on MoLE is pro- posed to enhance the multilingual and code-switching capabil- ities with less ...