This paper proposes a supervised learning method for detecting a semantic relation between a given pair of named entities, which may be located in different sentences. The method ...
—We present a novel framework to generate and rank plausible hypotheses for the spatial extent of objects in images using bottom-up computational processes and mid-level selectio...
With the goal to generate more scalable algorithms with higher efficiency and fewer open parameters, reinforcement learning (RL) has recently moved towards combining classical tec...
Abstract. Population size for EvolutionaryAlgorithms is usually an empirical parameter. We study the population size from aspects of fitness landscapes’ ruggedness and Probably ...
Motivated by the setting of reproducing kernel Hilbert space (RKHS) and its extensions considered in machine learning, we propose an RKHS framework for image and video colorizatio...