When we look at images taken from outdoor scenes, much of the information perceived is due to the ligthing conditions. In these scenes, the solar beams interact with the atmosphere...
This paper proposes a novel approach of combining an unsupervised clustering scheme called AutoClass with Hidden Markov Models (HMMs) to determine the traffic density state in a R...
Abstract. Recent results on robust density-based clustering have indicated that the uncertainty associated with the actual measurements can be exploited to locate objects that are ...
Abstract. Given an image, digital matting consists in extracting a foreground element from the background. Standard methods are initialized with a trimap, a partition of the image ...
Randomization is an economical and efficient approach for privacy preserving data mining (PPDM). In order to guarantee the performance of data mining and the protection of individ...