Sorry, you need to enable JavaScript to visit this website.

Smiling has a psychiatric effect in emotional state and may hold tremendous potential for clinical remediation in psychiatric disorders. A few researchers in image synthesis work on acting on the emotional state of subjects by automatically deforming their faces to synthesize joyful expression. However, to generate these expressions they apply the same deformation for the subjects while each person smiles differently. In this paper, we head towards a personalized synthesis of the joy expression.

The image-to-image translation task often associates with the problem of missing texture and edge information. In this paper, we proposed a framework to translate images while preserving more realistic textures and details. To this end, we disentangle the samples into shared content space and domain-specific style domain. Then, according to the blurred outlines and textures in the source domain, we introduce the classic canny edge detection algorithm to encode the boundary and edge information in the content latent space.