material for reviewers:

An approach to
real-time magnetic resonance imaging
for speech production

Submitted to JASA July/August 2003

Shrikanth Narayanan¹, Krishna Nayak^1,2, Sungbok Lee^1,3,
Abhinav Sethy¹, Dani Byrd³

USC Dept. of Electrical Engineering, Los Angeles, CA  90089-2564 USA;
²Stanford University Dept. of Electrical Engineering; ³USC Dept. of Linguistics

[Note: All the movies are played with the rate of 24 frames per second]

Figure 1:

[no movie]

Figure 2: "Say pea leap again"

Figure 3: "Say peal leap again"

Figure 4: "Say bean knee again"

Figure 5: Tamil retroflex [L]: "Andha vakyam paLam perusu."

Figure 6: Tamil [l] : "Andha vakyam palam perusu."

Figure 7: Tamil [zh]: "Andha vakyam pazham perusu."

Figure 8:

[no movie. Please refer to Fig. 8 in the paper]

Figure 9:: Image tracking demonstration