Dictionary learning from ambiguously labeled data

Yi Chen Chen, Vishal M. Patel, Jaishanker K. Pillai, Rama Chellappa, P. Jonathon Phillips

Research output: Contribution to journalConference articlepeer-review

20 Scopus citations


We propose a novel dictionary-based learning method for ambiguously labeled multiclass classification, where each training sample has multiple labels and only one of them is the correct label. The dictionary learning problem is solved using an iterative alternating algorithm. At each iteration of the algorithm, two alternating steps are performed: a confidence update and a dictionary update. The confidence of each sample is defined as the probability distribution on its ambiguous labels. The dictionaries are updated using either soft (EM-based) or hard decision rules. Extensive evaluations on existing datasets demonstrate that the proposed method performs significantly better than state-of-the-art ambiguously labeled learning approaches.

Original languageEnglish (US)
Article number6618896
Pages (from-to)353-360
Number of pages8
JournalProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
StatePublished - 2013
Externally publishedYes
Event26th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2013 - Portland, OR, United States
Duration: Jun 23 2013Jun 28 2013

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Vision and Pattern Recognition


  • Ambiguously labeled learning
  • dictionary-based learning


Dive into the research topics of 'Dictionary learning from ambiguously labeled data'. Together they form a unique fingerprint.

Cite this