Unsupervised manifold alignment using soft-assign technique

© 2016, Springer-Verlag Berlin Heidelberg. In this paper, we propose a robust unsupervised algorithm for automatic alignment of two manifolds in different datasets with possibly different dimensionalities. The significant contribution is that the proposed alignment algorithm is performed automatical...

Full description

Bibliographic Details
Main Authors: Fan, K., Mian, A., Liu, Wan-Quan, Li, L.
Format: Journal Article
Published: 2016
Online Access:http://hdl.handle.net/20.500.11937/40434
_version_ 1848755870549147648
author Fan, K.
Mian, A.
Liu, Wan-Quan
Li, L.
author_facet Fan, K.
Mian, A.
Liu, Wan-Quan
Li, L.
author_sort Fan, K.
building Curtin Institutional Repository
collection Online Access
description © 2016, Springer-Verlag Berlin Heidelberg. In this paper, we propose a robust unsupervised algorithm for automatic alignment of two manifolds in different datasets with possibly different dimensionalities. The significant contribution is that the proposed alignment algorithm is performed automatically without any assumptions on the correspondences between the two manifolds. For such purpose, we first automatically extract local feature histograms at each point of the manifolds and establish an initial similarity between the two datasets by matching their histogram-based features. Based on such similarity, an embedding space is estimated where the distance between the two manifolds is minimized while maximally retaining the original structure of the manifolds. The elegance of this idea is that such complicated problem is formulated as a generalized eigenvalue problem, which can be easily solved. The alignment process is achieved by iteratively increasing the sparsity of correspondence matrix until the two manifolds are correctly aligned and consequently one can reveal their joint structure. We demonstrate the effectiveness of our algorithm on different datasets by aligning protein structures, 3D face models and facial images of different subjects under pose and lighting variations. Finally, we also compare with a state-of-the-art algorithm and the results show the superiority of the proposed manifold alignment in terms of vision effect and numerical accuracy.
first_indexed 2025-11-14T09:03:10Z
format Journal Article
id curtin-20.500.11937-40434
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T09:03:10Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-404342017-09-13T13:38:56Z Unsupervised manifold alignment using soft-assign technique Fan, K. Mian, A. Liu, Wan-Quan Li, L. © 2016, Springer-Verlag Berlin Heidelberg. In this paper, we propose a robust unsupervised algorithm for automatic alignment of two manifolds in different datasets with possibly different dimensionalities. The significant contribution is that the proposed alignment algorithm is performed automatically without any assumptions on the correspondences between the two manifolds. For such purpose, we first automatically extract local feature histograms at each point of the manifolds and establish an initial similarity between the two datasets by matching their histogram-based features. Based on such similarity, an embedding space is estimated where the distance between the two manifolds is minimized while maximally retaining the original structure of the manifolds. The elegance of this idea is that such complicated problem is formulated as a generalized eigenvalue problem, which can be easily solved. The alignment process is achieved by iteratively increasing the sparsity of correspondence matrix until the two manifolds are correctly aligned and consequently one can reveal their joint structure. We demonstrate the effectiveness of our algorithm on different datasets by aligning protein structures, 3D face models and facial images of different subjects under pose and lighting variations. Finally, we also compare with a state-of-the-art algorithm and the results show the superiority of the proposed manifold alignment in terms of vision effect and numerical accuracy. 2016 Journal Article http://hdl.handle.net/20.500.11937/40434 10.1007/s00138-016-0772-8 restricted
spellingShingle Fan, K.
Mian, A.
Liu, Wan-Quan
Li, L.
Unsupervised manifold alignment using soft-assign technique
title Unsupervised manifold alignment using soft-assign technique
title_full Unsupervised manifold alignment using soft-assign technique
title_fullStr Unsupervised manifold alignment using soft-assign technique
title_full_unstemmed Unsupervised manifold alignment using soft-assign technique
title_short Unsupervised manifold alignment using soft-assign technique
title_sort unsupervised manifold alignment using soft-assign technique
url http://hdl.handle.net/20.500.11937/40434