Discriminative low-rank metric learning for face recognition

Lecture Notes in Computer Science, 2012

A proper distance metric is fundamental in many computer vision and pattern recognition applications such as classification, image retrieval, face recognition and so on. However, it is usually not clear what metric is appropriate for specific applications, therefore it becomes more reliable to learn a task oriented metric. Over the years, many metric learning approaches have been reported in literature. A typical one is to learn a Mahalanobis distance which is parameterized by a positive semidefinite (PSD) matrix M. An efficient method of estimating M is to treat M as a linear combination of rank-one matrices that can be learned using a boosting type approach. However, such approaches have two main drawbacks. First, the weight change across the training samples may be non-smooth. Second, the learned rank-one matrices might be redundant. In this paper, we propose a doubly regularized metric learning algorithm, termed by DRMetric, which imposes two regularizations on the conventional metric learning method. First, a regularization is applied on the weight of the training examples, which prevents unstable change of the weights and also prevents outlier examples from being weighed too much. Besides, a regularization is applied on the rank-one matrices to make them independent. This greatly reduces the redundancy of the rank-one matrices. We present experiments depicting the performance of the proposed method on a variety of datasets for various applications.

In this paper we study the problem of learning a low-rank (sparse) distance matrix. We propose a novel metric learning model which can simultaneously conduct dimension reduction and learn a distance matrix. The sparse representation involves a mixed-norm regularization which is non-convex. We then show that it can be equivalently formulated as a convex saddle (min-max) problem. From this saddle representation, we develop an efficient smooth optimization approach [17] for sparse metric learning, although the learning model is based on a nondifferentiable loss function. Finally, we run experiments to validate the effectiveness and efficiency of our sparse metric learning model on various datasets.

2007

In this paper, we propose a unified scheme of subspace and distance metric learning under the Bayesian framework for face recognition. According to the local distribution of data, we divide the k-nearest neighbors of each sample into the intra-person set and the inter-person set, and we aim to learn a distance metric in the embedding subspace, which can make the distances between the sample and its intra-person set smaller than the distances between it and its interperson set. To reach this goal, we define two variables, that is, the intra-person distance and the inter-person distance, which are from two different probabilistic distributions, and we model the goal with minimizing the overlap between two distributions. Inspired by the Bayesian classification error estimation, we formulate it by minimizing the Bhattachyrra coefficient between two distributions. The power of the proposed approach are demonstrated by a series of experiments on the CMU-PIE face database and the extended YALE face database.

Frontiers of Information Technology & Electronic Engineering, 2015

With the development of face recognition using sparse representation based classification (SRC), many relevant methods have been proposed and investigated. However, when the dictionary is large and the representation is sparse, only a small proportion of the elements contributes to the l 1-minimization. Under this observation, several approaches have been developed to carry out an efficient element selection procedure before SRC. In this paper, we employ a metric learning approach which helps find the active elements correctly by taking into account the interclass/intraclass relationship and manifold structure of face images. After the metric has been learned, a neighborhood graph is constructed in the projected space. A fast marching algorithm is used to rapidly select the subset from the graph, and SRC is implemented for classification. Experimental results show that our method achieves promising performance and significant efficiency enhancement.

Neurocomputing, 2015

Face recognition is one of the fundamental problems of computer vision and pattern recognition. Based on the recent success of Low-Rank Representation (LRR), we propose a novel image classification method for robust face recognition, named Low-Rank Representation-based Classification (LRRC). Based on seeking the lowest-rank representation of a set of test samples with respect to a set of training samples, the algorithm has the natural discrimination to perform classification. We also propose a Kernel Low-Rank Representation-based Classification (KLRRC), which is a nonlinear extension of LRRC. KLRRC is firstly utilized to face recognition. By using the kernel tricks, we implicitly map the input data into the kernel feature space associated with a kernel function. We construct a transformation matrix to reduce the dimensionality of the kernel feature space, where LRRC is performed. Experimental results on several face data sets demonstrate the effectiveness and robustness of our methods.

2015 IEEE International Conference on Computer Vision (ICCV), 2015

Sparse representation and low-rank matrix decomposition approaches have been successfully applied to several computer vision problems. They build a generative representation of the data, which often requires complex training as well as testing to be robust against data variations induced by nuisance factors. We introduce the invariant components, a discriminative representation invariant to nuisance factors, because it spans subspaces orthogonal to the space where nuisance factors are defined. This allows developing a framework based on geometry that ensures a uniform inter-class separation, and a very efficient and robust classification based on simple nearest neighbor. In addition, we show how the approach is equivalent to a local metric learning, where the local metrics (one for each class) are learned jointly, rather than independently, thus avoiding the risk of overfitting without the need for additional regularization. We evaluated the approach for face recognition with highly corrupted training and testing data, obtaining very promising results.

—We propose in this paper a parametric regularized locality preserving projections (LPP) method for face recognition. Our objective is to regulate the LPP space in a parametric manner and extract useful discriminant information from the whole feature space rather than a reduced projection subspace of principal component analysis. This results in better locality preserving power and higher recognition accuracy than the original LPP method. Moreover, the proposed regularization method can easily be extended to other manifold learning algorithms and to effectively address the small sample size problem. Experimental results on two widely used face databases demonstrate the efficacy of the proposed method. Index Terms—Face recognition, locality preserving projections (LPP), manifold learning, regularization, small sample size (SSS) problem.

2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014

This paper introduces a regularization method to explicitly control the rank of a learned symmetric positive semidefinite distance matrix in distance metric learning. To this end, we propose to incorporate in the objective function a linear regularization term that minimizes the k smallest eigenvalues of the distance matrix. It is equivalent to minimizing the trace of the product of the distance matrix with a matrix in the convex hull of rank-k projection matrices, called a Fantope. Based on this new regularization method, we derive an optimization scheme to efficiently learn the distance matrix. We demonstrate the effectiveness of the method on synthetic and challenging real datasets of face verification and image classification with relative attributes, on which our method outperforms state-of-the-art metric learning algorithms.

Procedia Computer Science, 2016

The sparse manifold clustering and embedding (SMCE) algorithm can be used to find a low-dimensional face manifold subspace. The neighborhoods are defined automatically and a sparse weight matrix is calculated with non zero values only for the members of the neighborhood. To obtain well separated classes in the low dimensional feature space, we propose to use discriminant functions during the embedding process. Existing manifold techniques do not give accurate results when images of different individuals with varying pose angles, lighting and facial expressions are considered. For these methods a proper choice of the neighborhood used to build the neighborhood graph is important. In SMCE this neighborhood is defined using a sparse optimization technique. K-nearest neighbor (KNN) classifier is used to classify the face images. The proposed method is compared with benchmarks using standard face datasets for varying facial expressions, lighting and poses. Leave-one-out cross validation testing strategy is used to validate the results.

2019 IEEE 21st International Workshop on Multimedia Signal Processing (MMSP)

In recent years, deep metric learning has achieved promising results in learning high dimensional semantic feature embeddings where the spatial relationships of the feature vectors match the visual similarities of the images. Similarity search for images is performed by determining the vectors with the smallest distances to a query vector. However, high retrieval quality does not depend on the actual distances of the feature vectors, but rather on the ranking order of the feature vectors from similar images. In this paper, we introduce a metric learning algorithm that focuses on identifying and modifying those feature vectors that most strongly affect the retrieval quality. We compute normalized approximated ranks and convert them to similarities by applying a nonlinear transfer function. These similarities are used in a newly proposed loss function that better contracts similar and disperses dissimilar samples. Experiments demonstrate significant improvement over existing deep feature embedding methods on the CUB-200-2011, Cars196, and Stanford Online Products data sets for all embedding sizes.