Riemannian Sparse Coding for Positive Definite Matrices

Anoop Cherian 1 Suvrit Sra 2
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Inspired by the great success of sparse coding for vector valued data, our goal is to represent symmetric positive definite (SPD) data matrices as sparse linear combinations of atoms from a dictionary, where each atom itself is an SPD matrix. Since SPD matrices follow a non-Euclidean (in fact a Riemannian) geometry, existing sparse coding techniques for Euclidean data cannot be directly extended. Prior works have approached this problem by defining a sparse coding loss function using either extrinsic similarity measures (such as the log-Euclidean distance) or kernelized variants of statistical measures (such as the Stein divergence, Jeffrey's divergence, etc.). In contrast, we propose to use the intrinsic Riemannian distance on the manifold of SPD matrices. Our main contribution is a novel mathematical model for sparse coding of SPD matrices; we also present a computationally simple algorithm for optimizing our model. Experiments on several computer vision datasets showcase superior classification and retrieval performance compared with state-of-the-art approaches.
Document type :
Conference papers
Complete list of metadatas

Cited literature [32 references]  Display  Hide  Download

Contributor : Thoth Team <>
Submitted on : Wednesday, December 3, 2014 - 4:56:06 PM
Last modification on : Monday, December 17, 2018 - 11:22:02 AM
Long-term archiving on : Monday, March 9, 2015 - 5:51:19 AM


Files produced by the author(s)




Anoop Cherian, Suvrit Sra. Riemannian Sparse Coding for Positive Definite Matrices. ECCV - European Conference on Computer Vision, Sep 2014, Zurich, Switzerland. pp.299-314, ⟨10.1007/978-3-319-10578-9_20⟩. ⟨hal-01057703⟩



Record views


Files downloads