There has been much work concentrated on creating accurate shot boundary detection algorithms in recent years. However a truly accurate method of cut detection still eludes researc...
Anthony Whitehead, Prosenjit Bose, Robert Lagani&e...
Abstract. In this paper, we suggest an algorithm to recover an image whose wavelet coefficients are partially lost. We propose a wavelet inpainting model by using L0-norm and the t...
Kernel k-means and spectral clustering have both been used to identify clusters that are non-linearly separable in input space. Despite significant research, these methods have re...
Many methods for object recognition, segmentation, etc., rely on tessellation of an image into "superpixels". A superpixel is an image patch which is better aligned with ...
—In this paper, we present a novel method to reconstruct the large scale scenes from multiple calibrated images. It first generates a quasi-dense 3D point cloud of the scene by m...