Mitigating Distribution Shifts in Pollen Classification from Microscopic Images Using Geometric Data Augmentations

Nam Cao, Olga Saukh

Research output: Contribution to conferencePaperpeer-review

Abstract

Distribution shifts are characterized by differences between the training and test data distributions. They can significantly reduce the accuracy of machine learning models deployed in real-world scenarios. This paper explores the distribution shift problem when classifying pollen grains from microscopic images collected in the wild with a low-cost camera sensor. We leverage the domain knowledge that geometric features are highly important for accurate pollen identification and introduce two novel geometric image augmentation techniques to significantly narrow the accuracy gap between the model performance on the train and test datasets. In particular, we show that Tenengrad and ImageToSketch filters are highly effective to balance the shape and texture information while leaving out unimportant details that may confuse the model. Extensive evaluations on various model architectures demonstrate a consistent improvement of the model generalization to field data of up to 14% achieved by the geometric augmentation techniques when compared to a wide range of standard image augmentations. The approach is validated through an ablation study using pollen hydration tests to recover the shape of dry pollen grains. The proposed geometric augmentations also receive the highest scores according to the affinity and diversity measures from the literature.
Original languageEnglish
Number of pages9
Publication statusPublished - 16 Dec 2023
Event29th IEEE International Conference on Parallel and Distributed Systems: IEEE ICPADS 2023 - Ocean Flower Island, Hainan, China
Duration: 17 Dec 202321 Dec 2023
https://ieee-cybermatics.org/2023/icpads/

Conference

Conference29th IEEE International Conference on Parallel and Distributed Systems
Abbreviated titleIEEE ICPADS
Country/TerritoryChina
CityHainan
Period17/12/2321/12/23
Internet address

Keywords

  • data augmentation
  • pollen classification

Fingerprint

Dive into the research topics of 'Mitigating Distribution Shifts in Pollen Classification from Microscopic Images Using Geometric Data Augmentations'. Together they form a unique fingerprint.

Cite this