SuperpixelGridMasks Data Augmentation: Application to Precision Health and Other Real-world Data - Archive ouverte HAL Access content directly
Journal Articles Journal of Healthcare Informatics Research Year : 2023

SuperpixelGridMasks Data Augmentation: Application to Precision Health and Other Real-world Data

Abstract

A novel approach of data augmentation based on irregular superpixel decomposition is proposed. This approach called SuperpixelGridMasks permits to extend original image datasets that are required by training stages of machine learning-related analysis architectures towards increasing their performances. Three variants named SuperpixelGridCut, SuperpixelGridMean, and SuperpixelGridMix are presented. These grid-based methods produce a new style of image transformations using the dropping and fusing of information. Extensive experiments using various image classification models as well as precision health and surrounding real-world datasets show that baseline performances can be significantly outperformed using our methods. The comparative study also shows that our methods can overpass the performances of other data augmentations. SuperpixelGridCut, SuperpixelGridMean, and SuperpixelGridMix codes are publicly available at https://github.com/hammoudiproject/SuperpixelGridMasks.
Fichier principal
Vignette du fichier
Hammoudi_SuperpixelGridMasks_Data_Augmentation_2204.08458.pdf (7.06 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03951542 , version 1 (24-01-2023)

Identifiers

Cite

Karim Hammoudi, Adnane Cabani, Bouthaina Slika, Halim Benhabiles, Fadi Dornaika, et al.. SuperpixelGridMasks Data Augmentation: Application to Precision Health and Other Real-world Data. Journal of Healthcare Informatics Research, 2023, 6, pp.442-460. ⟨10.1007/s41666-022-00122-1⟩. ⟨hal-03951542⟩
26 View
5 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More