Multivariate Analyses of Tongue Contours from Ultrasound Tongue Imaging

2026/02/28 by Stefano Coretta, Georges Sakr · 1 voice
Medicine · Psychology · Agricultural and Biological Sciences · #Traditional Chinese Medicine Studies #Phonetics and Phonology Research #Sensory Analysis and Statistical Methods

paper · doi:10.1177/00238309261419120

Abstract

This tutorial paper introduces two approaches to modeling tongue contour data obtained with DeepLabCut using multivariate generalized additive models (MGAMs) and multivariate functional principal component analysis (MFPCA). For each method, we present a fully commented analysis of two illustrative data sets: VC coarticulation in Italian and Polish, and consonant emphaticness in Lebanese Arabic. All the materials (inlcuding data and code) are available in the research compendium of the tutorial at https://github.com/stefanocoretta/mv_uti. We conclude by discussing advantages and disadvantages of the two methods (MGAM and MFPCA) and we recommend researchers to prefer MFPCA over MGAM as an initial step for modeling tongue contours.

Citations

• Hierarchical generalized additive models in ecology: an introduction with mgcv
• Analyzing dynamic phonetic data using generalized additive mixed modeling: A tutorial focusing on articulatory differences between L1 and L2 speakers of English

Discussions

• 🎉 NEW PAPER with @gsakr.bsky.social "Multivariate Analyses of Tongue Contours from Ultrasound Tongue Imaging" 👅📡🩻📉 doi.org/10.1177/0023... [bsky, 11 points, 1 comments]

Related

• Contour-based 3d tongue motion visualization using ultrasound image sequences
• Deep Learning for Automatic Tracking of Tongue Surface in Real-time Ultrasound Videos, Landmarks instead of Contours
• IrisNet: Deep Learning for Automatic and Real-time Tongue Contour Tracking in Ultrasound Video Data using Peripheral Vision
• Speech Map: A Statistical Multimodal Atlas of 4D Tongue Motion During Speech from Tagged and Cine MR Images
• A CNN-based tool for automatic tongue contour tracking in ultrasound images