Heilongjiang, China
Heilongjiang, China

Time filter

Source Type

Jin J.B.,Heilongjiang Institute of Finance | Liu Y.X.,Harbin Normal University
Applied Mechanics and Materials | Year: 2014

As an important field of image processing, image restoration has ever been a hot research topic. Our paper will focus on the adpative dictionary. We propose a novel improved adaptived dictionary learning algorithm. Utilizing the thoughts of manifold, for each atom in dictionry, we calculate the principal components of responding representaion errors as the tangent vector of the atom, which improves the representaion ability of dicitonary. Meanwhile, dictionary learned from the origin image is better at image denoising than adpative dictionary. © (2014) Trans Tech Publications, Switzerland.


Jin J.B.,Heilongjiang Institute of Finance | Liu Y.X.,Harbin Normal University
Applied Mechanics and Materials | Year: 2014

Representation and measurement are two important issues for saliency models. Different with previous works that learnt sparse features from large scale natural statistics, we propose to learn features from short-term statistics of single images. For saliency measurement, we defined basic firing rate (BFR) for each sparse feature, and then we propose to use feature activity rate (FAR) to measure the bottom-up visual saliency. The proposed FAR measure is biological plausible and easy to compute and with satisfied performance. Experiments on human trajectory positioning and psychological patterns demonstrate the effectiveness and robustness of our proposed method. © (2014) Trans Tech Publications, Switzerland.

Loading Heilongjiang Institute of Finance collaborators
Loading Heilongjiang Institute of Finance collaborators