Dictionary Learning from Ambiguously Labeled Data

Published

April 9, 2013

Author(s)

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

Abstract

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.

Proceedings Title

IEEE Conference on Computer Vision and Pattern Recognition

Conference Dates

June 25-27, 2013

Conference Location

Portland, OR

Pub Type

Conferences

Download Paper

Local Download

Citation

Phillips, P. , Chen, Y. , Patel, V. , Pillai, J. and Chellappa, R. (2013), Dictionary Learning from Ambiguously Labeled Data, IEEE Conference on Computer Vision and Pattern Recognition, Portland, OR, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=913138 (Accessed October 2, 2025)