Abstract:In a semi-supervised domain adaptation (DA) task, one has access to only few labeled target examples. In this case, the success of DA needs the effective utilization of a large number of unlabeled target data to extract more discriminative information that is useful for generalization. To this end, we exploit in this paper the feature space embeddings of the target data as well as multi-source prior models to augment the discrimination space for the target function learning. Therefore, we propose a novel multi-source adaptation learning framework based on Sparse Feature Space Representation (SFSR), or called SFSR-MSAL for short. Specifically, the SFSR algorithm is first presented for the further construction of robust graph, on which the discriminative information can be smoothly propagated into the unlabeled target data by additionally incorporating the geometric structure of the target data. Considering the robustness in the semi-supervised DA, we replace the traditional l 2-norm based least squares regression with the l 2, 1-norm sparse regression, and then construct the SFSR-graph based semi-supervised DA framework with multi-source adaptation constraints. Our framework is universal and can be easily degraded into semi-supervised learning by just tuning the regularization parameter. Moreover, to select the discriminative SFSR-graph Laplacians, we also introduce the ensemble SFSR-graph Laplacians regularization into SFSR-MSAL, thus further improving the performance of SFSR-MSAL. The validity of our methods including semi-supervised and DA learning are examined by several visual recognition tasks on some benchmark datasets, which demonstrate the superiority of our methods in comparison with other related state-of-the-art algorithms.

A General Approach for Achieving Supervised Subspace Learning in Sparse Representation

Image retrieval based on incremental subspace learning

Incremental semi-supervised subspace learning for image retrieval.

Automatic Subspace Learning via Principal Coefficients Embedding

Unsupervised Active Learning Via Subspace Learning

Unsupervised Subspace Learning Via Analysis Dictionary Learning.

Semi-supervised subspace learning with L2graph.

Learning dictionary via subspace segmentation for sparse representation

Orthogonal Principal Coefficients Embedding for Unsupervised Subspace Learning

A Supervised Subspace Learning Algorithm: Supervised Neighborhood Preserving Embedding

Unified Sparse Subspace Learning Via Self-Contained Regression

A linear subspace learning approach via sparse coding

Semi-supervised Domain Adaptation Via Subspace Exploration

Scene categorization via supervised subspace modeling and sparse representation

A generalized subspace projection approach for sparse representation classification

Unsupervised robust discriminative subspace representation based on discriminative approximate isometric embedding

Sparse feature space representation: A unified framework for semi-supervised and domain adaptation learning

Via subspace segmentation for sparse representation

Training Boosting-Like Algorithms with Semi-Supervised Subspace Learning

Towards Very Deep Representation Learning for Subspace Clustering

Learning Feature Sparse Principal Subspace.