In this paper a morphological tagging approach for document image invoice analysis is described. Tokens close by their morphology and confirmed in their location within different ...
We describe a clustering algorithm based on continuous Hidden Markov Models (HMM) to automatically classify both electrocardiogram (ECG) and intracranial pressure (ICP) beats base...
Cell segmentation is a challenging problem due to both the complex nature of the cells and the uncertainty present in video microscopy. Manual methods for this purpose are onerous...
The classical morphological segmentation paradigm is based on the watershed transform, constructed by flooding the gradient image seen as a topographic surface. For flooding a to...
Multi-word expressions constitute a significant portion of the lexicon of every natural language, and handling them correctly is mandatory for various NLP applications. Yet such e...