Description
This section displays the tongue, lips, and jaw pose from our data on a phone level basis. In our work, we used 39 phones without stress from CMUDict as we obtained the sequence of poses per phone by tagging the mocap sequences through the Montreal Forced Aligner. Then we obtained the mid pose from each phone's samples' sequences. Finally, we computed the mean pose per phone from all the mid poses.Visualization
In this section we visualize the frontal and sagittal views of each phones mean mid pose. The colored spheres represent the sensors, the red mesh represents the tongue, the palate's surface is an approximation from three sagittal and coronal traces. The lips and jaw are just displayed for reference and do not represent their actual size.AA
frontal view
sagittal view
AE
frontal view
sagittal view
AH
frontal view
sagittal view
AO
frontal view
sagittal view
AW
frontal view
sagittal view
AY
frontal view
sagittal view
B
frontal view
sagittal view
CH
frontal view
sagittal view
D
frontal view
sagittal view
DH
frontal view
sagittal view
EH
frontal view
sagittal view
ER
frontal view
sagittal view
EY
frontal view
sagittal view
F
frontal view
sagittal view
G
frontal view
sagittal view
HH
frontal view
sagittal view
IH
frontal view
sagittal view
IY
frontal view
sagittal view
JH
frontal view
sagittal view
K
frontal view
sagittal view
L
frontal view
sagittal view
M
frontal view
sagittal view
N
frontal view
sagittal view
NG
frontal view
sagittal view
OW
frontal view
sagittal view
OY
frontal view
sagittal view
P
frontal view
sagittal view
R
frontal view
sagittal view
S
frontal view
sagittal view
SH
frontal view
sagittal view
T
frontal view
sagittal view
TH
frontal view
sagittal view
UH
frontal view
sagittal view
UW
frontal view
sagittal view
V
frontal view
sagittal view
W
frontal view
sagittal view
Y
frontal view
sagittal view
Z
frontal view
sagittal view
ZH
frontal view
sagittal view