TY - JOUR
AU - Iqbal, A
AU - Seghouane, A-K
Y2 - 2020/02/09
Y2 - 2019/05/21
Y2 - 2019/05/21
Y2 - 2019/05/21
Y1 - 2019/11/01
SN - 1057-7149
UR - http://hdl.handle.net/11343/234154
AB - In this paper, a robust sequential dictionary learning (DL) algorithm is presented. The proposed algorithm is motivated from the maximum likelihood perspective on dictionary learning and its link to the minimization of the Kullback-Leibler divergence. It is obtained by using a robust loss function in the data fidelity term of the DL objective instead of the usual quadratic loss. The proposed robust loss function is derived from the α-divergence as an alternative to the Kullback-Leibler divergence, which leads to a quadratic loss. Compared to other robust approaches, the proposed loss has the advantage of belonging to class of redescending M-estimators, guaranteeing inference stability from large deviations from the Gaussian nominal noise model. The algorithm is obtained by solving a sequence of penalized rank-1 matrix approximation problems, where the ℓ 1 -norm is introduced as a penalty promoting sparsity and then using a block coordinate descent approach to estimate the unknowns. Performance comparison with similar robust DL algorithms on digit recognition, background removal, and gray-scale image denoising is performed highlighting efficacy of the proposed algorithm.
LA - English
PB - Institute of Electrical and Electronics Engineers
T1 - A alpha-Divergence-Based Approach for Robust Dictionary Learning
DO - 10.1109/TIP.2019.2922074
IS - IEEE Transactions on Image Processing
VL - 28
IS - 11
OP - FT130101394
L1 - /bitstream/handle/11343/234154/Iqbal_A%20alpha-Divergence.pdf?sequence=1&isAllowed=n
ER -