TY - GEN
T1 - Confusable Learning for Large-Class Few-Shot Classification
AU - Li, Bingcong
AU - Han, Bo
AU - Wang, Zhuowei
AU - Jiang, Jing
AU - Long, Guodong
N1 - Funding Information:
Preliminary work was done during an internship at Hong Kong Baptist University (HKBU). This research was partially funded by the Australian Government through the Australian Research Council (ARC) under grants LP180100654, HKBU Tier-1 Start-up Grant and HKBU CSD Start-up Grant.
Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021/2/25
Y1 - 2021/2/25
N2 - Few-shot image classification is challenging due to the lack of ample samples in each class. Such a challenge becomes even tougher when the number of classes is very large, i.e., the large-class few-shot scenario. In this novel scenario, existing approaches do not perform well because they ignore confusable classes, namely similar classes that are difficult to distinguish from each other. These classes carry more information. In this paper, we propose a biased learning paradigm called Confusable Learning, which focuses more on confusable classes. Our method can be applied to mainstream meta-learning algorithms. Specifically, our method maintains a dynamically updating confusion matrix, which analyzes confusable classes in the dataset. Such a confusion matrix helps meta learners to emphasize on confusable classes. Comprehensive experiments on Omniglot, Fungi, and ImageNet demonstrate the efficacy of our method over state-of-the-art baselines.
AB - Few-shot image classification is challenging due to the lack of ample samples in each class. Such a challenge becomes even tougher when the number of classes is very large, i.e., the large-class few-shot scenario. In this novel scenario, existing approaches do not perform well because they ignore confusable classes, namely similar classes that are difficult to distinguish from each other. These classes carry more information. In this paper, we propose a biased learning paradigm called Confusable Learning, which focuses more on confusable classes. Our method can be applied to mainstream meta-learning algorithms. Specifically, our method maintains a dynamically updating confusion matrix, which analyzes confusable classes in the dataset. Such a confusion matrix helps meta learners to emphasize on confusable classes. Comprehensive experiments on Omniglot, Fungi, and ImageNet demonstrate the efficacy of our method over state-of-the-art baselines.
KW - Confusion matrix
KW - Large-class few-shot classification
KW - Meta-learning
UR - http://www.scopus.com/inward/record.url?scp=85103273385&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-67661-2_42
DO - 10.1007/978-3-030-67661-2_42
M3 - Conference proceeding
AN - SCOPUS:85103273385
SN - 9783030676605
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 707
EP - 723
BT - Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2020, Proceedings
A2 - Hutter, Frank
A2 - Kersting, Kristian
A2 - Lijffijt, Jefrey
A2 - Valera, Isabel
PB - Springer Science and Business Media Deutschland GmbH
T2 - European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2020
Y2 - 14 September 2020 through 18 September 2020
ER -