Dictionary learning by Normalized Bilateral Projection

• Tezuka Taro

  University of Tsukuba

抄録

Dictionary learning is an unsupervised learning task that finds a set of template vectors that expresses input signals by sparse linear combinations. There are currently several methods for dictionary learning, for example K-SVD and MOD. In this paper, a new dictionary learning method, namely K-normalized bilateral projections (K-NBP), is proposed, which uses faster low rank approximation. Experiments showed that the method was fast and when the number of iterations was limited, it outperforms K-SVD. This indicated that the method was particularly suited to large data sets with high dimension, where each iteration takes a long time. K-NBP was applied to an image reconstruction task where images corrupted by noise were recovered using a dictionary learned from other images.

収録刊行物

• Journal of Information Processing

  Journal of Information Processing 24 (3), 565-572, 2016

  一般社団法人 情報処理学会

キーワード

• dictionary learning
• sparse coding
• bilateral projections
• image reconstruction

詳細情報

• CRID: 1390282680271625088
• NII論文ID: 130005151514
• NII論文ID: 170000147892
• NII書誌ID: AA11464847
• ISSN: 18827799
• ISSN: 18826652
• DOI: 10.2197/ipsjjip.24.565
• Web Site: http://id.nii.ac.jp/1001/00158412/
• Web Site: https://www.jstage.jst.go.jp/article/ipsjjip/24/3/24_565/_pdf
• 本文言語コード: en
• データソース種別:
  • JaLC
  • IRDB
  • Crossref
  • CiNii Articles
  • KAKEN
• 抄録ライセンスフラグ: 使用不可