Fig. 17From: Inner lips feature extraction based on CLNF with hybrid dynamic template for Cued Speech Y-axis represents the pixel number. Red curve: the error of the A parameter (expressed in pixels) between the values obtained by the CLNF and the ground truth value. Blue curve: the error of the A parameter (expressed in pixels) between the values obtained by the periodical spline method and the ground truth value. (a–c) correspond to the result of three speakersBack to article page