Dataset

We are collecting public plant disease datasets in PPDRD project.

Info of original plant related datasets

Name	env	Plant	img	class	Paper	Dataset
PlantVillage	lab	Multiple leaf	54,305	38	Paper	Dataset
PlantDocCls	internet	Multiple leaf	2,598	27	Paper	Dataset
TaiwanTomato	real+lab	Tomato leaf	622	5		Dataset
IVADLTomato	real	Tomato leaf	17,063	9		Dataset
IVADLRose	real	Rose leaf	23,114	6		Dataset
Apple2020	real	Apple leaf	3,642	4	Paper	Dataset
Apple2021	real	Apple leaf	18,632	6	Paper	Dataset
Cassava	real	Cassava leaf	21,397	5	Paper	Dataset
Citrus	lab	Citrus fruit leaf	105 + 609	5 + 5	Paper	Dataset
Rice5932	real	Rice leaf	5,932	4	Paper	Dataset
Rice1426	real	Rice leaf	1426	9	Paper	Dataset
CGIARWheat	real	wheat	876	3		Dataset
PDD271	real	Multiple leaf	220,592	271	Paper	Sample

Refer to visualize_dataset/dataset.md to see the detail info:

• 3 random images for each label
• the number of images for each label

Prepare the dataset for this project

You can download the original dataset used their links

To make the dataset for this project. After downloading the datasets: use ./data/make_*.py

To visualize the images for each datataset and each class

• use visualize_dataset/vis_dset.sh

PlantCLEF2022

• You can download the pretrained model here and see the paper

Train and test

• For CNN-based, normal case see /cnn_scripts/train.sh and /cnn_scripts/test.sh
• For CNN-based, few-shot case see /cnn_scripts/few_shot.sh and /cnn_scripts/test_few_shot.sh
• For ViT-based, see /scripts/train.sh and /scripts/test.sh
• For ViT-based, few-shot case see /scripts/few_shot.sh and /scripts/test_few_shot.sh

Cite our paper

• Published paper link
• BibTex:

@ARTICLE{xutransfer,
AUTHOR={Xu, Mingle and Yoon, Sook and Jeong, Yongchae and Park, Dong Sun},
TITLE={Transfer learning for versatile plant disease recognition with limited data},
JOURNAL={Frontiers in Plant Science},
VOLUME={13},
YEAR={2022},
URL={https://www.frontiersin.org/articles/10.3389/fpls.2022.1010981},
DOI={10.3389/fpls.2022.1010981},
ISSN={1664-462X},
}
@inproceedings{xu2022transfer,
  title={Transfer learning with self-supervised vision transformer for large-scale plant identification},
  author={Xu, Mingle and Yoon, Sook and Jeong, Yongchae and Lee, Jaesu and Park, Dong Sun},
  booktitle={International conference of the cross-language evaluation forum for European languages (Springer;)},
  pages={2253--2261},
  year={2022}
}
@article{xu2023plantclef2023,
  title={Plantclef2023: A bigger training dataset contributes more than advanced pretraining methods for plant identification},
  author={Xu, Mingle and Yoon, Sook and Wu, Chenmou and Baek, Jeonghyun and Park, Dong Sun},
  journal={Working Notes of CLEF},
  year={2023}
}