The problem of inferring 3D orientation of a camera from video sequences has been mostly addressed by first computing correspondences of image features. This intermediate step is ...
We present a method for animating 3D models of animals from existing live video sequences such as wild life documentaries. Videos are first segmented into binary images on which p...
The majority of the approaches to the automatic recovery of a panoramic image from a set of partial views are suboptimal in the sense that the input images are aligned, or registe...
An objective quality estimator for either still images or video should accurately estimate the perceived quality scores of a collection of stimuli. New applications and processing...
We introduce the first visual dataset of fast foods with a total of 4,545 still images, 606 stereo pairs, 303 3600 videos for structure from motion, and 27 privacy-preserving vide...