Detecting Spoof Voices in Asian Non-Native Speech: An Indonesian and Thai Case Study

\authorblockN Aulia Adila, Candy Olivia Mawalim, and Masashi Unoki \authorblockA Japan Advanced Institute of Science and Technology, Japan E-mail: {adila, candylim, unoki}@jaist.ac.jp · 2024

Read Detecting Spoof Voices in Asian Non-Native Speech: An Indonesian and Thai Case Study by \authorblockN Aulia Adila, Candy Olivia Mawalim, and Masashi Unoki \authorblockA Japan Advanced Institute of Science and Technology, Japan E-mail: {adila, candylim, unoki}@jaist.ac.jp (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-01040.txt.

Description

Aulia Adila, Candy Olivia Mawalim, and Masashi Unoki wrote this research paper, published in 2024. It studies how to spot fake or spoofed voices in speech from non-native speakers, using Indonesian and Thai speakers as the test case.

Voice spoofing can fool speaker checks, so detection systems matter. The team built a dataset with both native and non-native speech, pulled three kinds of sound features from it, and trained three standard classifiers. Detectors trained only on native speech struggled badly with non-native speech. Mixing non-native speech into the training data made detection much stronger. The paper argues that anti-spoofing tools must be trained on the accents they will actually meet, not just on native voices.

Catalog record

CREATOR\authorblockN Aulia Adila, Candy Olivia Mawalim, and Masashi Unoki \authorblockA Japan Advanced Institute of Science and Technology, Japan E-mail: {adila, candylim, unoki}@jaist.ac.jp
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-01040

Browse the full library · About the Tower of Babel