Background: A nanopore detector has a nanometer-scale trans-membrane channel across which a potential difference is established, resulting in an ionic current through the channel ...
Recent work shows how to use local spatio-temporal features to learn models of realistic human actions from video. However, existing methods typically rely on a predefined spatial...
Under the constraint of using only a single gallery image per person, this paper proposes a fast multi-class pattern classification approach to 2D face recognition robust to chan...
Shaokang Chen, Conrad Sanderson, Sai Sun, Brian C....
Abstract—Due to their ability to model sequential data without making unnecessary independence assumptions, conditional random fields (CRFs) have become an increasingly popular ...
For the transition from traditional to digital libraries, the large number of handwritten manuscripts that exist pose a great challenge. Easy access to such collections requires a...