Robust Human Matting via Semantic Guidance

Transductive Multi-Label Learning via Alpha Matting

Multi-label learning deals with the problems when each instance can be assigned to multiple classes simultaneously, which are ubiquitous in real-world learning tasks. In this paper, we propose a new multilabel learning method, which is able to exploit unlabeled data to obtain an effective model for assigning appropriate multiple labels to instances. The proposed method is called T (TRansduct...

Image Matting via LLE/iLLE Manifold Learning

Accurately extracting foreground objects is the problem of isolating the foreground in images and video, called image matting which has wide applications in digital photography. This problem is severely ill-posed in the sense that, at each pixel, one must estimate the foreground and background pixels and the so-called alpha value from only pixel information. The most recent work in natural imag...

Semantic guidance with SCOTT

Proof Planning with Semantic Guidance

Proof planning [2, 10] is a reasoning process in which mathematical theorem proving is viewed as a planning problem. It incorporates domain-dependent and/or general mathematical knowledge, and shows a great potential in the area where the traditional logicbased theorem proving systems have been typically weak. The plan operator – method – encodes a piece of mathematical knowledge. When there ar...

Energy Matting

In this paper, we formulate the image matting as one of solving energy minimization problems. Our approach has the following advantages. First, the matte estimation is modeled using an energy function as a MRF optimization problem. Second, the energy function combines the gradient of the matte, the gradient of the color and statistical sampling together to achieve global optimization. Third, th...