The approach for predictive analysis focuses on a business strategy to streamline business-customer relationships in order to maximize client satisfaction and thereby improve cust...
Traditional classification methods assume that the training and the test data arise from the same underlying distribution. However, in several adversarial settings, the test set is...
Spectral analysis approaches have been actively studied in machine learning and data mining areas, due to their generality, efficiency, and rich theoretical foundations. As a natur...
Dijun Luo, Heng Huang, Chris H. Q. Ding, Feiping N...
We recently introduced the watershed cuts, a notion of watershed in edge-weighted graphs. In this paper, our main contribution is a thinning paradigm from which we derive three alg...
Jean Cousty, Gilles Bertrand, Laurent Najman, Mich...
We propose an efficient orthonormal wavelet-domain video denoising algorithm based on an appropriate integration of motion compensation into an adapted version of our recently devi...