Abstract:In transfer learning model, the source domain samples and target domain samples usually share the same class labels but have different distributions. In general, the existing transfer learning algorithms ignore the interclass differences and intraclass similarities across domains. To address these problems, this article proposes a transfer learning algorithm based on discriminative Fisher embedding and adaptive maximum mean discrepancy (AMMD) constraints, called discriminative Fisher embedding dictionary transfer learning (DFEDTL). First, combining the label information of source domain and part of target domain, we construct the discriminative Fisher embedding model to preserve the interclass differences and intraclass similarities of training samples in transfer learning. Second, an AMMD model is constructed using atoms and profiles, which can adaptively minimize the distribution differences between source domain and target domain. The proposed method has three advantages: 1) using the Fisher criterion, we construct the discriminative Fisher embedding model between source domain samples and target domain samples, which encourages the samples from the same class to have similar coding coefficients; 2) instead of using the training samples to design the maximum mean discrepancy (MMD), we construct the AMMD model based on the relationship between the dictionary atoms and profiles; thus, the source domain samples can be adaptive to the target domain samples; and 3) the dictionary learning is based on the combination of source and target samples which can avoid the classification error caused by the difference among samples and reduce the tedious and expensive data annotation. A large number of experiments on five public image classification datasets show that the proposed method obtains better classification performance than some state-of-the-art dictionary and transfer learning methods. The code has been available at https://github.com/shilinrui/DFEDTL.

Learning The Discriminative Dictionary For Sparse Representation By A General Fisher Regularized Model

Learning Individual-Specific Dictionaries With Fused Multiple Features For Face Recognition

Learning Discriminative Dictionary for Group Sparse Representation.

A Sparse Representation Model Using the Complete Marginal Fisher Analysis Framework and Its Applications to Visual Recognition

Discriminative Dictionary Learning Via Fisher Discrimination K-SVD Algorithm

Toward Efficient Image Representation: Sparse Concept Discriminant Matrix Factorization

A New Fisher Discriminative K-Svd Algorithm For Face Recognition

Face Recognition by Sparse Discriminant Analysis Via Joint L2,1-norm Minimization

Local Feature Discriminant Projection

Laplacian Sparse Dictionary Learning for Image Classification Based on Sparse Representation

Face Recognition - a Generalized Marginal Fisher Analysis Approach.

Joint Structured Sparsity Regularized Multiview Dimension Reduction for Video-Based Facial Expression Recognition.

Joint Subspace Recovery and Enhanced Locality Driven Robust Flexible Discriminative Dictionary Learning

Discriminative Fisher Embedding Dictionary Transfer Learning for Object Recognition

Sparse Concept Discriminant Matrix Factorization for Image Representation.

Block-Diagonal Sparse Representation by Learning a Linear Combination Dictionary for Recognition

Learning Multi-Task Sparse Representation Based on Fisher Information

Generalized robust linear discriminant analysis for jointly sparse learning

A Generalized Form of Fisher Linear Discriminant Function

Locality-regularized Linear Regression Discriminant Analysis for Feature Extraction

Face Recognition Based on Dictionary Learning and Subspace Learning