Article: Visualisation of Chinese phonemes based on three-dimensional tongue model and ultrasound images Journal: International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC) 2026 Vol.51 No.5 pp.1 - 11 Abstract: This study introduces a high-precision 3D tongue motion visualisation model using ultrasound imaging to aid pronunciation training for hearing-impaired individuals and Chinese learners. Unlike static 2D images or generic 3D representations, this approach reconstructs authentic tongue motion by integrating 2D ultrasound data with a high-fidelity 3D model. It provides side-by-side 2D/3D comparisons for commonly confused phonemes and evaluates accuracy using curve-similarity metrics instead of point-error measures. Results show 91.7% reconstruction accuracy, capturing subtle dynamic tongue movements during pronunciation. The model dynamically and accurately visualises articulation, offering a more precise tool for pronunciation training in special education and second language acquisition. Inderscience Publishers - linking academia, business and industry through research

Title: Visualisation of Chinese phonemes based on three-dimensional tongue model and ultrasound images

Authors: Shaochuan Zhang; Yihuai Zhang; Ping He

Addresses: Beihang University, Beijing, 100083, China ' Beihang University, Beijing, 100083, China ' Beihang University, Beijing, 100083, China

Abstract: This study introduces a high-precision 3D tongue motion visualisation model using ultrasound imaging to aid pronunciation training for hearing-impaired individuals and Chinese learners. Unlike static 2D images or generic 3D representations, this approach reconstructs authentic tongue motion by integrating 2D ultrasound data with a high-fidelity 3D model. It provides side-by-side 2D/3D comparisons for commonly confused phonemes and evaluates accuracy using curve-similarity metrics instead of point-error measures. Results show 91.7% reconstruction accuracy, capturing subtle dynamic tongue movements during pronunciation. The model dynamically and accurately visualises articulation, offering a more precise tool for pronunciation training in special education and second language acquisition.

Keywords: tongue motion visualisation; ultrasound imaging; pronunciation training; curve-similarity; statistic model.

DOI: 10.1504/IJAHUC.2026.152017

International Journal of Ad Hoc and Ubiquitous Computing, 2026 Vol.51 No.5, pp.1 - 11

Received: 28 Jul 2025
Accepted: 10 Jan 2026
Published online: 02 Mar 2026 *

Title: Visualisation of Chinese phonemes based on three-dimensional tongue model and ultrasound images

Keep up-to-date