A. Dutta, X. Li, P. Richtarik
Index Terms Wighted low-rank approximation, 1-norm minimization, Robust PCA, Background modelling
We primarily study a special a weighted low-rank pproximation of matrices and then apply it to solve the ackground modeling problem. We propose two algorithms for his purpose: one operates in the batch mode on the entire data and the other one operates in the batch-incremental mode on he data and naturally captures more background variations and omputationally more effective. Moreover, we propose a robust echnique that learns the background frame indices from the ata and does not require any training frames. We demonstrate through extensive experiments that by inserting a simple weight in the Frobenius norm, it can be made robust to the outliers similar to theorm. Our methods match or outperform several tate-of-the-art online and batch background modeling methodsin virtually all quantitative and qualitative measures.