Understanding Matrix Function Normalizations in Covariance Pooling through the Lens of Riemannian Geometry

• URL: http://arxiv.org/abs/2407.10484v2
• Date: Sat, 20 Jul 2024 08:11:10 GMT
• Title: Understanding Matrix Function Normalizations in Covariance Pooling through the Lens of Riemannian Geometry
• Authors: Ziheng Chen, Yue Song, Xiao-Jun Wu, Gaowen Liu, Nicu Sebe,
• Abstract summary: Global Covariance Pooling (GCP) has been demonstrated to improve the performance of Deep Neural Networks (DNNs) by exploiting second-order statistics of high-level representations.
• Score: 63.694184882697435
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Global Covariance Pooling (GCP) has been demonstrated to improve the performance of Deep Neural Networks (DNNs) by exploiting second-order statistics of high-level representations. GCP typically performs classification of the covariance matrices by applying matrix function normalization, such as matrix logarithm or power, followed by a Euclidean classifier. However, covariance matrices inherently lie in a Riemannian manifold, known as the Symmetric Positive Definite (SPD) manifold. The current literature does not provide a satisfactory explanation of why Euclidean classifiers can be applied directly to Riemannian features after the normalization of the matrix power. To mitigate this gap, this paper provides a comprehensive and unified understanding of the matrix logarithm and power from a Riemannian geometry perspective. The underlying mechanism of matrix functions in GCP is interpreted from two perspectives: one based on tangent classifiers (Euclidean classifiers on the tangent space) and the other based on Riemannian classifiers. Via theoretical analysis and empirical validation through extensive experiments on fine-grained and large-scale visual classification datasets, we conclude that the working mechanism of the matrix functions should be attributed to the Riemannian classifiers they implicitly respect.

Related papers

• Optimal Matrix-Mimetic Tensor Algebras via Variable Projection [0.0]
  Matrix mimeticity arises from interpreting tensors as operators that can be multiplied, factorized, and analyzed analogous to matrices. We learn optimal linear mappings and corresponding tensor representations without relying on prior knowledge of the data. We provide original theory of uniqueness of the transformation and convergence analysis of our variable-projection-based algorithm.
  arXiv  Detail & Related papers  (2024-06-11T04:52:23Z)
• Synergistic eigenanalysis of covariance and Hessian matrices for enhanced binary classification [75.90957645766676]
  We present a novel approach that combines the eigenanalysis of a covariance matrix evaluated on a training set with a Hessian matrix evaluated on a deep learning model. Our approach is substantiated by formal proofs that establish its capability to maximize between-class mean distance and minimize within-class variances.
  arXiv  Detail & Related papers  (2024-02-14T16:10:42Z)
• Learning Graphical Factor Models with Riemannian Optimization [70.13748170371889]
  This paper proposes a flexible algorithmic framework for graph learning under low-rank structural constraints. The problem is expressed as penalized maximum likelihood estimation of an elliptical distribution. We leverage geometries of positive definite matrices and positive semi-definite matrices of fixed rank that are well suited to elliptical models.
  arXiv  Detail & Related papers  (2022-10-21T13:19:45Z)
• Semi-Supervised Subspace Clustering via Tensor Low-Rank Representation [64.49871502193477]
  We propose a novel semi-supervised subspace clustering method, which is able to simultaneously augment the initial supervisory information and construct a discriminative affinity matrix. Comprehensive experimental results on six commonly-used benchmark datasets demonstrate the superiority of our method over state-of-the-art methods.
  arXiv  Detail & Related papers  (2022-05-21T01:47:17Z)
• Riemannian statistics meets random matrix theory: towards learning from high-dimensional covariance matrices [2.352645870795664]
  This paper shows that there seems to exist no practical method of computing the normalising factors associated with Riemannian Gaussian distributions on spaces of high-dimensional covariance matrices. It is shown that this missing method comes from an unexpected new connection with random matrix theory. Numerical experiments are conducted which demonstrate how this new approximation can unlock the difficulties which have impeded applications to real-world datasets.
  arXiv  Detail & Related papers  (2022-03-01T03:16:50Z)
• Robust Geometric Metric Learning [17.855338784378]
  This paper proposes new algorithms for the metric learning problem. A general approach, called Robust Geometric Metric Learning (RGML), is then studied. The performance of RGML is asserted on real datasets.
  arXiv  Detail & Related papers  (2022-02-23T14:55:08Z)
• Automatic differentiation for Riemannian optimization on low-rank matrix and tensor-train manifolds [71.94111815357064]
  In scientific computing and machine learning applications, matrices and more general multidimensional arrays (tensors) can often be approximated with the help of low-rank decompositions. One of the popular tools for finding the low-rank approximations is to use the Riemannian optimization.
  arXiv  Detail & Related papers  (2021-03-27T19:56:00Z)
• Probabilistic Learning Vector Quantization on Manifold of Symmetric Positive Definite Matrices [3.727361969017079]
  We develop a new classification method for manifold-valued data in the framework of probabilistic learning vector quantization. In this paper, we generalize the probabilistic learning vector quantization algorithm for data points living on the manifold of symmetric positive definite matrices. Empirical investigations on synthetic data, image data, and motor imagery EEG data demonstrate the superior performance of the proposed method.
  arXiv  Detail & Related papers  (2021-02-01T06:58:39Z)
• Understanding Implicit Regularization in Over-Parameterized Single Index Model [55.41685740015095]
  We design regularization-free algorithms for the high-dimensional single index model. We provide theoretical guarantees for the induced implicit regularization phenomenon.
  arXiv  Detail & Related papers  (2020-07-16T13:27:47Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.