This is tensorflow implementation for paper "Deep Image Matting".
Thanks to Davi Frossard, "vgg16_weights.npz" can be found his blog: "https://www.cs.toronto.edu/~frossard/post/vgg16/"
2017-8-25: Now this code can be used to train, but the data is owned by company. I'll try my best to provide code and model that can do inference. And current code still has some problems like slow preprocess, huge memory occupation etc....