Speech-Guided Multimodal Learning for Vocal Tract Segmentation in Real-Time MRI

TL;DR

This paper introduces a novel multimodal learning framework that uses acoustic and phonological data during training to improve real-time MRI vocal tract segmentation, enabling accurate inference without audio.

Contribution

The proposed three-stage framework effectively leverages multimodal supervision to enhance vocal tract segmentation in real-time MRI, even when only imaging data is available during inference.

Findings

Outperforms existing unimodal and multimodal methods on benchmark datasets.

Demonstrates transfer of multimodal knowledge into single-modality inference.

Provides a clinically deployable solution for vocal tract segmentation.

Abstract

Segmenting vocal tract articulators in real-time MRI (rtMRI) is a challenging dynamic image segmentation problem characterized by low contrast, rapid motion, and limited spatial resolution. However, while rtMRI acquisitions may provide synchronized acoustic signals, existing methods discard this information, and the few multimodal approaches that incorporate audio cannot be deployed when audio is unavailable. We propose a three-stage framework that leverages acoustic and phonological supervision during training while requiring only the rtMRI image at inference: phonological representations are converted into spatial bounding-box priors for articulator localization, visual and acoustic encoders are aligned via dual-level cross-modal contrastive pretraining, and the learned representations are fused through a cross-attention decoder, effectively transferring multimodal knowledge into a single-modality inference pipeline. Evaluated on 75-Speaker~Annot-16 and USC-TIMIT datasets, our method outperforms existing unimodal and multimodal methods, demonstrating that multimodal supervision provides transferable benefits for precise and clinically deployable vocal tract segmentation.