Automatic image matting and fusing for portrait synthesis

Yi Z K, Song W F, Li S, et al

Sci China Inf Sci, 2022, 65(2): 124101

We propose an automatic image matting and fusing system for portrait synthesis in this study. We firstly use a face detection algorithm to determine if the input contains a face. Then, we use a semantic segmentation neural network to generate a trimap and feed the trimap and the portrait into the neural network to predict the alpha channel value. Finally, the input portrait's background is replaced with the given background via an image synthesis algorithm to obtain the synthesized portrait.

PDF Slides