作者位次 #11/17ACL 2026Poster· Natural Language Processing
S pid R -Adapt: A Universal Speech Representation Model for Few-Shot Adaptation
Mahi Luthra · Jiayi Shen · Maxime Poli · Angelo Ortiz Tandazo · Yosuke Higuchi · Youssef Benchekroun · Martin Gleize · Charles-Éric Saint-James · Dongyan Lin · Phillip Rust · Angel Villar · Surya Parimi · Vanessa Stark · Rashel Moritz · Juan Pino · Yann LeCun · Emmanuel Dupoux
摘要
Human infants, with only a few hundred hours of speech exposure, acquire basic units of new languages, highlighting a striking efficiency gap compared to the data-hungry self-supervised speech models. To address this gap, this paper introduces SpidR-Adapt for rapid adaptation of speech units to n...