xml94 / ppdrd

Collect public plant disease recognition datasets

PPDRD

New contribution is encouraged and appreciated.

Collect public plant disease recognition datasets.

This project aims to only collect the public dataset to recognize plant disease because the community can not verify the performance on the private ones, although they have information and contributions.

• For every dataset, a file is linked to give more description.
• Dataset name: we will give a name if a dataset has no name. Default is fully public and PartPublic means partially public.
• Crop: show the crop if only one or give the number of crops, otherwise.
• Number of classes: include diseased classes and healthy if having.
• Number of images: Only those images with public labels are counted, and only the original images are counted (augmented images are not).
• Image background: complex (cmpx), medium (med), simple (simp).
• Machine learning (ML) task: image classification (clf), object detection (obj), segmentation (seg).
• Performance (PE): official leaderboard for challenges, or reported results in the dataset publication.
• Reference: default is with publications or official challenges such as in kaggle; otherwise no reference .

Dataset name	Crop	Class	Image	Image BG	ML task & PE
Apple2020	Apple	4	1,821	med	clf: 0.984 AUROC
Apple2021	Apple	6	18,632	med	clf: 0.883 F1
PCApple2023	Apple	9	10,212	med+sim	clf: N.A
ASDID	Soybean	8	9,648	med+sim	clf: 0.968 Acc
BRACOL	Coffee	5	1,747	sim	clf: 0.956 Acc
RoCoLe	Coffee	6	1,560	med	clf: N.A
iCassava	Cassava	5	5,656	med	clf: 0.939 Acc
CLDCMakerere	Cassava	5	21,397	cmpx+med	clf: 0.913 Acc
CLDCAmanda	Cassava	6	2,249	med	clf: 0.930 Acc
CLDD	Cassava	3	228	med	clf: N.A
CDRD	Cucumber	8	1,289	med+sim	clf: N.A
CucumberNegm	Cucumber	2	691	med	clf: N.A
PaddyDoctor	Rice	10	10,407	cmpx	clf: 0.990 Acc
Rice1426	Rice	9	1,426	cmpx+med+sim	clf: 0.971 Acc
Rice5932	Rice	4	5,932	med	clf: 0.984 Acc
HuyDoRice	Rice	4	3,355	sim	clf: 0.984 Acc
DhanShomadhan	Rice	5	1,106	cmpx+sim	clf: N.A
WheatLong	Wheat	5	999	cmpx	clf: 0.971 Acc
WheatLeafDataset	Wheat	3	407	med+sim	clf: N.A
GroundNutLeaf	Groundnut	5	3,058	med	clf: N.A
MaizeCraze	Corn	6	2,355	sim	clf: N.A
BisqueCorn	Corn	2	1,785	cmpx	clf: N.A
CornNLB	Corn	1	18,222	cmpx	clf: N.A
iBean	Bean	3	1,296	med	clf: N.A
SoybeanMignoni	Soybean	3	6,410	cmpx	clf: N.A
TaiwanTomato	Tomato	6	622	med+sim	clf: N.A
GLFD	Guava	5	527	sim	clf: N.A
IDADP	Grape	7	3,596	cmpx	clf: N.A
CitrusRauf	Citrus	10	759	sim	clf: N.A
PlantVillage	14	38	54,305	sim	clf: N.A
FieldPV	14	38	665	med+sim	clf: 0.720 Acc
PlantDocCls	13	27	2,598	cmpx+med+sim	clf: N.A
PlantConservation	12	10	4,503	sim	clf: N.A
CCMT	4	22	24,881	med+sim	clf: N.A
PDD271	N.A	271	2,710	cmpx+med	clf: 0.855 Acc
PlantDocObj	13	27	2,598	cmpx+med+sim	obj: N.A
NZDLPlantDiseaseV1	5	20	3,337	med	obj: 0.745 mAP
NZDLPlantDiseaseV2	8	28	3,039	med	obj: 0.932 mAP
FieldPlant	4	31	5,156	cmpx+med	obj: 0.144 mAP
GrapevineDiseaseMalo	Grape	3	744	cmpx	obj: N.A
GrapevineDiseaseMalo	Grape	4	128	cmpx	seg: N.A
RoCoLe	Coffee	2	1,560	sim	seg: N.A
ATLDSD	Apple	5	1,641	med+sim	seg: N.A

Reference

Please consider cite our related papers if you think this project is useful.

@article{xu2023plant,
  title={Plant Disease Recognition Datasets in the Age of Deep Learning: Challenges and Opportunities},
  author={Xu, Mingle and Park, Ji Eun and Lee, Jaehwan and Yang, Jucheng and Yoon, Sook},
  journal={arXiv preprint arXiv:2312.07905},
  year={2023}
}
@article{meng2023known,
  title={Known and unknown class recognition on plant species and diseases},
  author={Meng, Yao and Xu, Mingle and Kim, Hyongsuk and Yoon, Sook and Jeong, Yongchae and Park, Dong Sun},
  journal={Computers and Electronics in Agriculture},
  volume={215},
  pages={108408},
  year={2023},
  publisher={Elsevier}
}
@article{xu2023embracing,
  title={Embracing limited and imperfect training datasets: opportunities and challenges in plant disease recognition using deep learning},
  author={Xu, Mingle and Kim, Hyongsuk and Yang, Jucheng and Fuentes, Alvaro and Meng, Yao and Yoon, Sook and Kim, Taehyun and Park, Dong Sun},
  journal={Frontiers in Plant Science},
  volume={14},
  year={2023},
  publisher={Frontiers Media SA}
}
@article{xu2022transfer,
  title={Transfer learning for versatile plant disease recognition with limited data},
  author={Xu, Mingle and Yoon, Sook and Jeong, Yongchae and Park, Dong Sun},
  journal={Frontiers in Plant Science},
  volume={13},
  pages={1010981},
  year={2022},
  publisher={Frontiers}
}