The IMage Euclidean Distance (IMED) is a class of image metrics, in which the spatial relationship between pixels is taken into consideration. It was shown that calculating the IM...
We introduce a linearly weighted variant of the total
variation for vector fields in order to formulate regularizers
for multi-class labeling problems with non-trivial interclass...
The ability to construct CAD or other object models from edge and range data has a fundamental meaning in building a recognition and positioning system. While the problem of model...
Embedding algorithms search for low dimensional structure in complex data, but most algorithms only handle objects of a single type for which pairwise distances are specified. Thi...
Amir Globerson, Gal Chechik, Fernando C. Pereira, ...
In this paper, we introduce a new approach for content-based similarity search for brain images. Based on the keyblock representation, our framework employs the Principal Componen...