G-IFT: A Gated Linear Unit adapter with Iterative Fine-Tuning for Low-Resource Children’s Speaker Verification

Vishwas, Jiusi, Abeer · 2025

Read G-IFT: A Gated Linear Unit adapter with Iterative Fine-Tuning for Low-Resource Children’s Speaker Verification by Vishwas, Jiusi, Abeer (2025) 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-2508-07836.txt.

Description

Vishwas Shetty, Jiusi Zheng, and Abeer Alwan adapt adult speaker verification to children's voices in a 2025 study. Systems trained on grown-up speech stumble on children because young voices differ acoustically, and recordings of children are too scarce for ordinary fine-tuning to fix the gap.

Their framework inserts a gated adapter into a pretrained speaker embedding model, then fine-tunes in repeated rounds. Gating lets the adapter pass or reshape features selectively, improving transfer from the data-rich adult domain. Better child speaker verification supports classroom tools and safe voice interfaces for young users, where weak checks carry real risk.

Catalog record

CREATORVishwas, Jiusi, Abeer
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE21 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-07836

Browse the full library · About the Tower of Babel