Research
Language Discrimination Improves Linguistic Learning in Multilingual Speech Models
Multilingual self-supervised speech models can benefit from sharing information across languages, but under a matched total pretraining data budget they still fall short of monolingual models. We show
Multilingual self-supervised speech models can benefit from sharing information across languages, but under a matched total pretraining data budget they still fall short of monolingual models. We show that strengthening the model’s ability to discriminate languages during pretraining reduces and, on some measures, closes this multilingual gap on continuous phonetic and higher-level linguistic measures, while preserving substantial cross-language sharing. Using a controlled English/French HuBERT setting, we test two interventions which strengthen language discrimination: an auxiliary language…
Related
- BabyHuBERT: Multilingual Self-Supervised Learning for Segmenting Speakers in Child-Centered Long-Form Recordings
- Layer-wise Probing of wav2vec 2.0 and Whisper for Consonant Cluster Reduction in African American English
- An Exploration of Mamba for Speech Self-Supervised Models
Source: Apple ML Research | 2026-10-02