Fig. 18
From: Inner lips feature extraction based on CLNF with hybrid dynamic template for Cued Speech
The figures are plotted in the parameter A-B plane with all the vowels in one repetition for the first subject. a CLNF with corrected B parameter but non corrected A parameter for the round lips. b CLNF with both corrected B and A parameters for the third viseme with automatic detection of the third viseme; stars correspond to the first viseme, circles to the second viseme, and triangles to the third viseme. The color order is blue, red, green, magenta, and cyan. They are corresponding with the vowel order shown in Table 1