A new algorithm is described for refining the pose of a model of a rigid object, to conform more accurately to the image structure. Elemental 3D forces are considered to act on the...
Anthony D. Worrall, Geoffrey D. Sullivan, Keith D....
Semantic region labeling in outdoor scenes, e.g., identifying sky, grass, foliage, water, and snow, facilitates content-based image retrieval, organization, and enhancement. A maj...
Matthew R. Boutell, Jiebo Luo, Christopher M. Brow...
Virtually all methods in image processing and computer vision, for removing weather effects from images, assume single scattering of light by particles in the atmosphere. In reali...
Many computer vision algorithms limit their performance by ignoring the underlying 3D geometric structure in the image. We show that we can estimate the coarse geometric propertie...
This paper describes the continuing development of a system for tracking multiple man made objects, (typically vehicles) moving in a natural open world scene, where the detected m...