The bounding box representation employed by many popular object detection models [3, 6] implicitly assumes all pixels inside the box belong to the object. This assumption makes th...
In this paper we reformulate the 3D reconstruction of deformable surfaces from monocular video sequences as a labeling problem. We solve simultaneously for the assignment of featu...
We propose to detect abnormal events via a sparse reconstruction over the normal bases. Given an over-complete normal basis set (e.g., an image sequence or a collection of local s...
Abstract. Image matting and segmentation are two closely related topics that concern extracting the foreground and background of an image. While the methods based on global optimiz...
In this paper, we proposed a fast and accurate human pose estimation framework that combines top-down and bottom-up methods. The framework consists of an initialization stage and a...
Yanchao Su, Haizhou Ai, Takayoshi Yamashita, Shiho...