Most existing methods of active support vector machine (ASVM) focus on the samples nearby the current separating hyperplane, which ignore some support vector (SV) samples which are far form the separating hyperplane, also pay not attention on whether the current separating hyperplane is close to the optimal one. In this paper a new classification method of ASVM based on improved probability-calculation method is presented. It not only presents a new method of calculating probability, but also measures the degree of closeness of the current separating hyperplane to the actual separating hyperplane by a confidence factor. Experimental results show the superiority of our proposed method both in classification accuracy and computing cost.
In this paper, we present a novel no-reference blur metric for images. The blur metric is based on analyzing image features include the mean value of phase congruency image, the entropy of phase congruency image and the distorted image, and the gradient of the distorted image. The new index does NOT need any information from reference image, and image quality estimation is accomplished by simple functional relationship between those features. Our experimental results show that the new index outperforms existing popular no-reference blurriness metric and full reference PSNR on LIVE Gaussian blurred database and IVC blurring images.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.