Abstract:Distance metric learning has been widely studied in multifarious research fields. The mainstream approaches learn a Mahalanobis metric or learn a linear transformation. Recent related works propose learning a linear combination of base vectors to approximate the metric. In this way, fewer variables need to be determined, which is efficient when facing high-dimensional data. Nevertheless, such works obtain base vectors using additional data from related domains or randomly generate base vectors. However, obtaining base vectors from related domains requires extra time and additional data, and random vectors introduce randomness into the learning process, which requires sufficient random vectors to ensure the stability of the algorithm. Moreover, the random vectors cannot capture the rich information of the training data, leading to a degradation in performance. Considering these drawbacks, we propose a novel distance metric learning approach by introducing base vectors explicitly learned from training data. Given a specific task, we can make a sparse approximation of its objective function using the top eigenvalues and corresponding eigenvectors of a predefined integral operator on the reproducing kernel Hilbert space. Because the process of generating eigenvectors simply refers to the training data of the considered task, our proposed method does not require additional data and can reflect the intrinsic information of the input features. Furthermore, the explicitly learned eigenvectors do not result in randomness, and we can extend our method to any kernel space without changing the objective function. We only need to learn the coefficients of these eigenvectors, and the only hyperparameter that we need to determine is the number of eigenvectors that we utilize. Additionally, an optimization algorithm is proposed to efficiently solve this problem. Extensive experiments conducted on several datasets demonstrate the effectiveness of our proposed method.

Image classification algorithm based on elliptic hyperbolic mahalanobis metric

Face Recognition Based on Local Binary Pattern and Improved Pairwise-constrained Multiple Metric Learning

Discriminative Cnn Via Metric Learning For Hyperspectral Classification

Bilevel Distance Metric Learning for Robust Image Recognition.

The Spectral Matching Algorithm Based on Hyperbolic Mahalanobis Metric

Eigenvector-Based Distance Metric Learning for Image Classification and Retrieval.

Hyperspectral Imagery Classification with Deep Metric Learning

A Mahalanobis Metric Learning-Based Polynomial Kernel for Classification of Hyperspectral Images

Image Matching Algorithm Based on Mahalanobis-Distance Spectral Features

LAM3L: Locally Adaptive Maximum Margin Metric Learning for Visual Data Classification

Multidimensional Local Binary Pattern for Hyperspectral Image Classification

Exponential Discriminative Locality Alignment for Hyperspectral Image Classification

A Novel Center-Boundary Metric Loss to Learn Discriminative Features for Hyperspectral Image Classification

Local Correntropy Matrix Representation for Hyperspectral Image Classification

Local Image Distance Metric Learning

A scalable algorithm for learning a mahalanobis distance metric

Multiple Characteristics Similarity Metric Method for Hyperspectral Image Classification

Classification Using Global and Local Mahalanobis Distances

Probabilistic Deep Metric Learning for Hyperspectral Image Classification

A Dual-Layer Supervised Mahalanobis Kernel for the Classification of Hyperspectral Images

Learning Pseudo Metric for Intelligent Multimedia Data Classification and Retrieval