Computer vision tasks such as learning, recognition, classification or segmentation applied to spatial data often requires spatial normalization of repeated features and structure...
In China-US Million Book Digital Library, output of the digitalization process is more than one terabyte of text in OEB and PDF format. To access these data quickly and accurately,...
—This paper presents a robust, low-cost ADC code hit counting technique to record the number of times each ADC output code word appears with respect to the ramp input. Using a sm...
The direct visual inspection of WCE video by an expert is a tiring and cost activity and it is a true bottleneck to the widespread application of this diagnostic technique. In this...
Giovanni Gallo, Eliana Granata, Giuseppe Scarpulla
We describe a new tagging model where the states of a hidden Markov model (HMM) estimated by unsupervised learning are incorporated as the features in a maximum entropy model. Our...