Existing line drawing techniques that are commonly used to depict shapes and structures lack a sense of spatial arrangement, especially for complex scenes. In this paper we utilize depth information...

Disocclusion that appears when synthesizing virtual views from RGB-D scenes is performed through the integration of depth into the classical patch-based algorithm of Criminisi. The proposed...

We propose an exemplar-based video completion algorithm and geometry-guided space-time artifact reduction technique. The completion algorithm extends a method working on still images, to videos. The...

A method that imperceptibly embeds binary images into the gradient domain of a host CIELAB color image, by modifying the color gradient vectors under a JND constraint. It also supports to hide...

We build a semantic vector representation of illustrations by training CNNs. As the proposed vector space correctly reflects the semantic meanings of illustrations, users can efficiently search...

We propose a novel representation method called Content-Aware Cube Unwrapping for panoramas based on orientational rectification, energy estimation, image unwrapping and seam carving. Thus, many...

In this paper, we propose an acceleration method for image denoising with redundant discrete cosine transform (R-DCT). Experimental results show that our method can accelerate R-DCT, while the...

In this paper, we present a novel example-based method for creating a painted image from a photograph. Our method allows us to establish automatically a new painted image preserving the features of...