GenKL: An Iterative Framework for Resolving Label Ambiguity and Label Non-conformity in Web Images Via a New Generalized KL Divergence
Open Access
- 15 July 2023
- journal article
- research article
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 131 (11) , 3035-3059
Abstract
Web image datasets curated online inherently contain ambiguous in-distribution instances and out-of-distribution instances, which we collectively callnon-conforming(NC) instances. In many recent approaches for mitigating the negative effects of NC instances, the core implicit assumption is that the NC instances can be found via entropy maximization. For “entropy” to be well-defined, we are interpreting the output prediction vector of an instance as the parameter vector of a multinomial random variable, with respect to some trained model with a softmax output layer. Hence, entropy maximization is based on the idealized assumption that NC instances have predictions that are “almost” uniformly distributed. However, in real-world web image datasets, there are numerous NC instances whose predictions are far from being uniformly distributed. To tackle the limitation of entropy maximization, we propose $$(\alpha , \beta )$$ -generalized KL divergence, $${\mathcal {D}}_{\text {KL}}^{\alpha , \beta }(p\Vert q)$$ , which can be used to identify significantly more NC instances. Theoretical properties of $${\mathcal {D}}_{\text {KL}}^{\alpha , \beta }(p\Vert q)$$ are proven, and we also show empirically that a simple use of $${\mathcal {D}}_{\text {KL}}^{\alpha , \beta }(p\Vert q)$$ outperforms all baselines on the NC instance identification task. Building upon $$(\alpha ,\beta )$$ -generalized KL divergence, we also introduce a new iterative training framework,GenKL, that identifies and relabels NC instances. When evaluated on three web image datasets, Clothing1M, Food101/Food101N, and mini WebVision 1.0, we achieved new state-of-the-art classification accuracies: $$81.34\%$$ , $$85.73\%$$ and $$78.99\%$$ / $$92.54\%$$ (top-1/top-5), respectively.
Keywords
All Related Versions
Funding Information
- National Research Foundation Singapore (AISG-RP-2019-015)
This publication has 43 references indexed in Scilit:
- CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images2018
- Deep Bilevel Learning2018
- Delta Divergence: A Novel Decision Cognizant Measure of Classifier IncongruenceIEEE Transactions on Cybernetics, 2018
- Making Deep Neural Networks Robust to Label Noise: A Loss Correction Approach2017
- A decision cognizant Kullback–Leibler divergencePattern Recognition, 2017
- Deep Residual Learning for Image Recognition2016
- Food-101 – Mining Discriminative Components with Random Forests2014
- The kappa statistic as a function of sensitivity and specificityInternational Journal of Mathematical Education in Science and Technology, 2005
- Some interlacing properties of the Schur complement of a Hermitian matrixLinear Algebra and its Applications, 1992
- Divergence measures based on the Shannon entropyIEEE Transactions on Information Theory, 1991