arXiv subjects

Michele Gubian

Publications and source records attributed to Michele Gubian.

3 recordsLinked to original sources

arXiv2026

Perceptual compensation for tonal context in self-supervised speech models

This study examines the extent to which the wav2vec2.0 architecture exhibits evidence of compensation for phonological context. We conducted a pseudo-replication of a perceptional compensation experiment on Mandarin Chinese tones, and compared the embedding similarities and probing classifier outputs between a purely self-supervised pre-trained model and a model fine-tuned for Mandarin ASR. No evidence of compensation was found in the embedding similarities of the purely pre-trained model. Probing classifiers showed some evidence of compensation in addition to the expected layer-wise improvements in categorization, but failed to replicate human performance on isolated test syllables. Our findings contrast with previous reports of sensitivity to phonological structure emerging through pre-training alone, and suggest that supervised objectives may be necessary to encourage the abstraction of at least some types of phonological regularities.

cs.CL↗

arXiv2025

Analyzing the relationships between pretraining language, phonetic, tonal, and speaker information in self-supervised speech models

Analyses of self-supervised speech models have begun to reveal where and how they represent different types of information. However, almost all analyses have focused on English. Here, we examine how wav2vec2 models trained on four different languages encode both language-matched and non-matched speech. We use probing classifiers and geometric analyses to examine how phones, lexical tones, and speaker information are represented. We show that for all pretraining and test languages, the subspaces encoding phones, tones, and speakers are largely orthogonal, and that layerwise patterns of probing accuracy are similar, with a relatively small advantage for matched-language phone and tone (but not speaker) probes in the later layers. Our findings suggest that the structure of representations learned by wav2vec2 is largely independent of the speech material used during pretraining.

cs.CL↗

arXiv2024

Statistics in Phonetics

Phonetics is the scientific field concerned with the study of how speech is produced, heard and perceived. It abounds with data, such as acoustic speech recordings, neuroimaging data, or articulatory data. In this paper, we provide an introduction to different areas of phonetics (acoustic phonetics, sociophonetics, speech perception, articulatory phonetics, speech inversion, sound change, and speech technology), an overview of the statistical methods for analyzing their data, and an introduction to the signal processing methods commonly applied to speech recordings. A major transition in the statistical modeling of phonetic data has been the shift from fixed effects to random effects regression models, the modeling of curve data (for instance via GAMMs or FDA methods), and the use of Bayesian methods. This shift has been driven in part by the increased focus on large speech corpora in phonetics, which has been driven by machine learning methods such as forced alignment. We conclude by identifying opportunities for future research.

stat.AP↗