Automatic image matting and fusing for portrait synthesis  

作　　者：Zhike YI Wenfeng SONG Shuai LI Aimin HAO 

机构地区：[1]State Key Laboratory of Virtual Reality Technology and Systems,Beihang University,Beijing 100191,China

出　　处：《Science China(Information Sciences)》2022年第2期231-233,共3页中国科学（信息科学）（英文版）

基　　金：supported in part by National Key R&D Program of China (Grant No. 2018YFB1700603);National Natural Science Foundation of China (Grant Nos. 61672077, 61532002);Beijing Natural Science Foundation-Haidian Primitive Innovation Joint Fund (Grant No. L182016)。

摘　　要：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.

关 键 词：NEURAL IMAGE SYNTHESIS 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]