Image Reconstruction by Sparse Coding and Selective Attention

Sparse coding theory demonstrates that the neurons in primary visual cortex form a sparse representation of natural scenes in the viewpoint of statistics. In this paper, inspired by the research of selective attention in psychology, we propose a novel self-adaptive algorithm to further reduce the activated variables. The experimental results show that the quality of reconstructed images obtained by our method is satisfying. Moreover, combining selective attention and sparse coding our method evidently decreases the number of coefficients which may be activated and preserves the main information at the same time.