Advancements in Multilingual Speech Technologies for Human-Computer Interaction
- 22 October 2025
- book chapter
- Published by IGI Global Scientific Publishing
- p. 101-132
Abstract
The global integration of communication networks increases the demand for multilingual human-computer interaction (HCI) systems. Speech technology innovations that support multiple languages enhance accessibility and usability. This chapter explores how automatic speech recognition (ASR) and text-to-speech (TTS), combined with language models, ena-ble real-time speech translation across diverse accents and linguistic systems. Deep learning and transfer learning help models learn from small, complex language datasets, even under noisy conditions. Practical applications include educational tools, virtual assistants, cus-tomer service, and healthcare. The section also addresses the need for culturally intuitive, unbiased systems through inclusive design and ethical guidelines, promoting linguistic equality. Advances in multilingual speech technology make HCI more inclusive and effec-tive for global users, supporting developers and researchers alike.Keywords
This publication has 35 references indexed in Scilit:
- Immigrant linguistic justice: The lay of the landMetaphilosophy, 2023
- Speech Act Theory and Ethics of Speech Processing as Distinct Stages: the ethics of collecting, contextualizing and the releasing of (speech) data2023
- Improving Luxembourgish Speech Recognition with Cross-Lingual Speech Representations2023
- Racial disparities in automated speech recognitionProceedings of the National Academy of Sciences, 2020
- Building Inclusive Language Classroom Spaces through Multilingual Writing Practices for Newly-Arrived Students in NorwayScandinavian Journal of Educational Research, 2019
- Voice and Speech Synthesis—Highlighting the Control of Prosody2018
- JSpeech: A Multi-Lingual Conversational Speech Corpus2018
- Machine translation based data augmentation for Cantonese keyword spotting2016
- Prosody Control and Variation Enhancement Techniques for HMM-Based Expressive Speech Synthesis2015
- Reliable Pitch Marking of Affective Speech at Peaks or Valleys Using Restricted Dynamic ProgrammingIEEE Transactions on Multimedia, 2010