Using gaussian processes to improve zero-shot learning with relative attributes

Yeshi Dolma, Vinay P. Namboodiri

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Relative attributes can serve as a very useful method for zero-shot learning of images. This was shown by the work of Parikh and Grauman [1] where an image is expressed in terms of attributes that are relatively specified between different class pairs. However, for zero-shot learning the authors had assumed a simple Gaussian Mixture Model (GMM) that used the GMM based clustering to obtain the label for an unknown target test example. In this paper, we contribute a principled approach that uses Gaussian Process based classification to obtain the posterior probability for each sample of an unknown target class, in terms of Gaussian process classification and regression for nearest sample images. We analyse different variants of this approach and show that such a principled approach yields improved performance and a better understanding in terms of probabilistic estimates. The method is evaluated on standard Pubfig and Shoes with Attributes benchmarks.

Original languageEnglish
Title of host publicationComputer Vision - 13th Asian Conference on Computer Vision, ACCV 2016, Revised Selected Papers
EditorsKo Nishino, Shang-Hong Lai, Vincent Lepetit, Yoichi Sato
PublisherSpringer Verlag
Pages150-164
Number of pages15
ISBN (Print)9783319541921
DOIs
Publication statusPublished - 1 Jan 2017
Event13th Asian Conference on Computer Vision, ACCV 2016 - Taipei, Taiwan, Province of China
Duration: 20 Nov 201624 Nov 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10115 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th Asian Conference on Computer Vision, ACCV 2016
CountryTaiwan, Province of China
City Taipei
Period20/11/1624/11/16

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Dolma, Y., & Namboodiri, V. P. (2017). Using gaussian processes to improve zero-shot learning with relative attributes. In K. Nishino, S-H. Lai, V. Lepetit, & Y. Sato (Eds.), Computer Vision - 13th Asian Conference on Computer Vision, ACCV 2016, Revised Selected Papers (pp. 150-164). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10115 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-54193-8_10