Published August 17, 2022 | Version 0.1.0

cropped diBas dataset for machine learning

Authors/Creators

Description

This zip file contains a collection of 244 x 244 images of various gram-stained bacteria strains, cropped from the diBas dataset (Bartosz Zieli ́nski et al. 2017).  Each folder is named by the species of the bacteria images it contains. 

Potential uses for the dataset include machine learning. An example neural network for bacteria classification using this dataset is deployed at https://huggingface.co/spaces/qile0317/Bacteria-Classification

Note that the data is not completely optimized, as it contains trace amounts of almost blank images. 

Notes

To further improve viability in machine learning applications, a future version will be implemented to incorporate varying degrees of tint (warmth), size, and warping of the images. Of course, image augmentation of the original data via basic image manipulation code is also possible.

Files

cropped_DiBas.zip

Files (2.2 GB)

Name Size
md5:e89ead4a80a46865cb4a33d5144120b4
2.2 GB Preview Download

Additional details

References

  • Bartosz Zieli ́nski et al. "Deep learning approach to bacterial colony classification". In: PloS One 12.9 (2017), e0184554