Sparse representaion via #x2113;1/2-norm minimization for facial expression recognition

摘要

The ℓ1/2-norm regularizer is shown to have many promising properties such as unbiasedness, sparsity and oracle properties. By exploiting these properties of ℓ1/2-norm regularizer, we propose a novel model of sparse representation based classification via ℓ1/2-norm minimization (ℓ1/2-SRC) for facial expression recognition in this paper. In ℓ1/2-SRC, we use ℓ1/2-norm minimization as an alternative to ℓ0-norm minimization instead of using ℓ1-norm minimization in the traditional ℓ1-SRC. By adopting ℓ1/2-norm minimization, we can find a sparser and more accurate solution than the ℓ1-SRC, and the optimization problem of ℓ1/2-norm minimization is much easier to be solved than that of ℓ0-norm minimization. Furthermore, an active-set based iterative reweighted algorithm is proposed to solve the ℓ1/2-norm minimization problem. The experimental results on JAFFE and Cohn-Kanade databases testify the efficiency of ℓ1/2-SRC.