14thAnnual Conference of the International Speech Communication Association

Lyon, France
August 25-29, 2013

Vocal Tract Cross-Distance Estimation from Real-Time MRI Using Region-of-Interest Analysis

Adam Lammert (1), Vikram Ramanarayanan (1), Michael Proctor (2), Shrikanth Narayanan (1)

(1) University of Southern California, USA
(2) University of Western Sydney, Australia

Real-Time Magnetic Resonance Imaging affords speech articulation data with good spatial and temporal resolution and complete midsagittal views of the moving vocal tract, but also brings many challenges in the domain of image processing and analysis. Region-of-interest analysis has previously been proposed for simple, efficient and robust extraction of linguistically-meaningful constriction degree information. However, the accuracy of such methods has not been rigorously evaluated, and no method has been proposed to calibrate the pixel intensity values or convert them into absolute measurements of length. This work provides such an evaluation, as well as insights into the placement of regions in the image plane and calibration of the resultant pixel intensity measurements. Measurement errors are shown to be generally at or below the spatial resolution of the imaging protocol with a high degree of consistency across time and overall vocal tract configuration, validating the utility of this method of image analysis.

Full Paper

Bibliographic reference.  Lammert, Adam / Ramanarayanan, Vikram / Proctor, Michael / Narayanan, Shrikanth (2013): "Vocal tract cross-distance estimation from real-time MRI using region-of-interest analysis", In INTERSPEECH-2013, 959-962.