We present an approach to low-level vision that combines two main ideas: the use of convolutional networks as an image processing architecture and an unsupervised learning procedu...
We consider the problem of learning classifiers in structured domains, where some objects have a subset of features that are inherently absent due to complex relationships between...
Gal Chechik, Geremy Heitz, Gal Elidan, Pieter Abbe...
We introduce a fully automatic algorithm which optimizes the high-level structure of a given quadrilateral mesh to achieve a coarser quadrangular base complex. Such a topological ...
Abstract— Wireless sensor networks consist of small, inexpensive devices which interact with the environment, communicate with each other, and perform distributed computations in...
Manish Kushwaha, Isaac Amundson, Xenofon D. Koutso...
The majority of existing pedestrian trackers concentrate on maintaining the identities of targets, however systems for remote biometric analysis or activity recognition in surveill...