Sparse feature selection based on L2,1/2-matrix norm for web image annotation

• Spare Feature Selection method based on l2,1/2-matrix norm is proposed.
• Shared subspace learning is exploited to consider the feature correlation.
• A effective algorithm for optimizing the objective function is introduced.
• The convergence of the algorithm is proven.
• Experiments demonstrate the method is suit for web image annotation.

Web image annotation based on sparse feature selection has received an increasing amount of interest in recent years. However, existing sparse feature selection methods become less effective and efficient. This raises an urgent need to develop good sparse feature selection methods to improve web image annotation performance. In this paper we propose a novel sparse feature selection framework for web image annotation, namely Sparse Feature Selection based on L2,1/2-matrix norm (SFSL). SFSL can select more sparse and more discriminative features by exploiting the l2,1/2-matrix norm with shared subspace learning, and then improve the web image annotation performance. We proposed an efficient iterative algorithm to optimize the objective function. Extensive experiments are performed on two web image datasets. The experimental results have validated that our method outperforms the state-of-the-art algorithms and suits for large-scale web image annotation.