Book Chapter

Learning with Biased Complementary Labels

X Yu, T Liu, M Gong, D Tao

Computer Vision – ECCV 2018 | Lecture Notes in Computer Science | Published : 2018

Abstract

In this paper, we study the classification problem in which we have access to easily obtainable surrogate for true labels, namely complementary labels, which specify classes that observations do not belong to. Let Y and Y be the true and complementary labels, respectively. We first model the annotation of complementary labels via transition probabilities P(Y=i|Y=j), i\ne j\in \{1,\cdots,c\}, where c is the number of classes. Previous methods implicitly assume that P(Y=i|Y=j), \forall i\ne j, are identical, which is not true in practice because humans are biased toward their own experience. For example, as shown in Fig. 1, if an annotator is more familiar with monkeys than prairie dogs when p..

View full abstract

University of Melbourne Researchers

Grants

Citation metrics