… But with deep learning based matting, the matting module is highly nonlinear and … the
semantic information from input hardly retains. To combine the coarse semantics and ne matting …

… semantic human matting without trimaps as extra input. Specifically, we train a mask prediction
network to estimate the coarse semantic … A matting refinement network takes in the unified …

… human matting framework, named Semantic Guided Human Matting (SGHM). It builds on
a … As mentioned above, we propose to improve matting results by using semantic human …

… We introduce semantics into the matting task and demonstrate how semantic information
can be used to achieve the most competitive performance. To our knowledge, this is the first …

… This paper extends the general semantic matting method to domain specific matting. Given
a foreground object of a certain class, its alpha matte can exhibit multiple alpha matte patterns…

… Semantic human matting aims to accurately separate the humans from a given image . It is a
… Given an RGB image as input, human matting can be divided into two subtasks: foreground …

… To capture both the local and global semantic context features from the input image, we
adopt a Deeplab [6] like encoder-decoder architecture as shown in Figure 2. Moreover, inspired …

… These factors make human matting more challenging. Therefore, in this work, we focus on
human matting, using segmentation networks as the basis to extract rough contours of human …

… semantic correspondences of all human instances to obtain a united semantics guidance,
which discriminates human … We collect a large-scale human instance matting dataset, named …

… human matting more challenging. Therefore, in this work, we focus on the human matting …
to assist in the detail extraction of the human contour. By adding it to MODNet, we get a new …