Entropy-Based Uncertainty Calibration for Generalized Zero-Shot Learning
Paper
Paper/Presentation Title | Entropy-Based Uncertainty Calibration for Generalized Zero-Shot Learning |
---|---|
Presentation Type | Paper |
Authors | Chen, Zhi, Huang, Zi, Li, Jingjing and Zhang, Zheng |
Journal or Proceedings Title | Proceedings of the 32nd Australasian Database Conference (ADC 2021) |
Journal Citation | 12610, pp. 139-151 |
Number of Pages | 13 |
Year | 2021 |
Publisher | Springer |
Place of Publication | Switzerland |
ISSN | 1611-3349 |
0302-9743 | |
ISBN | 9783030693763 |
9783030693770 | |
Digital Object Identifier (DOI) | https://doi.org/10.1007/978-3-030-69377-0_12 |
Web Address (URL) of Paper | https://link.springer.com/chapter/10.1007/978-3-030-69377-0_12 |
Web Address (URL) of Conference Proceedings | https://link.springer.com/book/10.1007/978-3-030-69377-0 |
Conference/Event | 32nd Australasian Database Conference (ADC 2021) |
Event Details | 32nd Australasian Database Conference (ADC 2021) Parent Australasian Database Conference Delivery In person Event Date 29 Jan 2021 to end of 05 Feb 2021 Event Location Dunedin, New Zealand Rank B B B B B B |
Abstract | Compared to conventional zero-shot learning (ZSL) where recognising unseen classes is the primary or only aim, the goal of generalized zero-shot learning (GZSL) is to recognise both seen and unseen classes. Most GZSL methods typically learn to synthesise visual representations from semantic information on the unseen classes. However, these types of models are prone to overfitting the seen classes, resulting in distribution overlap between the generated features of the seen and unseen classes. The overlapping region is filled with uncertainty as the model struggles to determine whether a test case from within the overlap is seen or unseen. Further, these generative methods suffer in scenarios with sparse training samples. The models struggle to learn the distribution of high dimensional visual features and, therefore, fail to capture the most discriminative inter-class features. To address these issues, in this paper, we propose a novel framework that leverages dual variational autoencoders with a triplet loss to learn discriminative latent features and applies the entropy-based calibration to minimize the uncertainty in the overlapped area between the seen and unseen classes. To calibrate the uncertainty for seen classes, we calculate the entropy over the softmax probability distribution from a general classifier. With this approach, recognising the seen samples within the seen classes is relatively straightforward, and there is less risk that a seen sample will be misclassified into an unseen class in the overlapped region. Extensive experiments on six benchmark datasets demonstrate that the proposed method outperforms state-of-the-art approaches. |
Keywords | Generalized zero shot learning; Image classification; Transfer learning; Triplet network |
Contains Sensitive Content | Does not contain sensitive content |
ANZSRC Field of Research 2020 | 4602. Artificial intelligence |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Series | Lecture Notes in Computer Science |
Byline Affiliations | University of Queensland |
University of Electronic Science and Technology of China, China | |
Harbin Institute of Technology, China |
https://research.usq.edu.au/item/zyx1x/entropy-based-uncertainty-calibration-for-generalized-zero-shot-learning
9
total views0
total downloads5
views this month0
downloads this month