Complex Object Recognition using Deep Deformable Regularized Sparse Learning

Accurate object recognition in complex transformation and multi-modal is a challenging problem in computer vision tasks. An efficient complex object recognition method using Deep Deformable Regularized Sparse Learning (DDRSL) is proposed, which not only can handle the deformable and multi-modal problems, but also can leverage datasets, features, attributes, classes and their high-level semantic descriptions to learn about their relations with unsupervised transfer learning. Extensive qualitative and quantitative experiments are performed, and the results show the proposed method outperforms the several state-of-the-art methods.