Scalable deep k-subspace clustering
Date
2019
Authors
Zhang, T.
Ji, P.
Harandi, M.
Hartley, R.
Reid, I.
Editors
Jawahar, C.V.
Li, H.
Mori, G.
Schindler, K.
Li, H.
Mori, G.
Schindler, K.
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Lecture Notes in Artificial Intelligence, 2019 / Jawahar, C.V., Li, H., Mori, G., Schindler, K. (ed./s), vol.11365, pp.466-481
Statement of Responsibility
Tong Zhang, Pan Ji, Mehrtash Harandi, Richard Hartley, and Ian Reid
Conference Name
Asian Conference on Computer Vision (ACCV) (2 Dec 2018 - 6 Dec 2018 : Perth, Australia)
Abstract
Subspace clustering algorithms are notorious for their scalability issues because building and processing large affinity matrices are demanding. In this paper, we introduce a method that simultaneously learns an embedding space along subspaces within it to minimize a notion of reconstruction error, thus addressing the problem of subspace clustering in an end-to-end learning paradigm. To achieve our goal, we propose a scheme to update subspaces within a deep neural network. This in turn frees us from the need of having an affinity matrix to perform clustering. Unlike previous attempts, our method can easily scale up to large datasets, making it unique in the context of unsupervised learning with deep architectures. Our experiments show that our method significantly improves the clustering accuracy while enjoying cheaper memory footprints.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
© Springer Nature Switzerland AG 2019