Investigating and Explaining the Frequency Bias in Image Classification

Zhiyu Lin; Yifei Gao; Jitao Sang

IJCAI 2022poster15 citations

Investigating and Explaining the Frequency Bias in Image Classification

Abstract

CNNs exhibit many behaviors different from humans, one of which is the capability of employing high-frequency components. This paper discusses the frequency bias phenomenon in image classification tasks: the high-frequency components are actually much less exploited than the low- and mid- frequency components. We first investigate the frequency bias phenomenon by presenting two observations on feature discrimination and learning priority. Furthermore, we hypothesize that (1) the spectral density, (2) class consistency directly affect the frequency bias. Specifically, our investigations verify that the spectral density of datasets mainly affects the learning priority, while the class consistency mainly affects the feature discrimination.

AI Ethics, Trust, Fairness: Trustworthy AIAI Ethics, Trust, Fairness: BiasAI Ethics, Trust, Fairness: Explainability and InterpretabilityAI Ethics, Trust, Fairness: Safety & RobustnessComputer Vision: Bias, Fairness & Privacy

BibTeX

@inproceedings{ijcai2022p101,
  title     = {Investigating and Explaining the Frequency Bias in Image Classification},
  author    = {Lin, Zhiyu and Gao, Yifei and Sang, Jitao},
  booktitle = {Proceedings of the Thirty-First International Joint Conference on
               Artificial Intelligence, {IJCAI-22}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Lud De Raedt},
  pages     = {717--723},
  year      = {2022},
  month     = {7},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2022/101},
  url       = {https://doi.org/10.24963/ijcai.2022/101},
}