Back to datasets
Dataset assetOpen Source CommunityPlant Disease IdentificationMultimodal Data Analysis

PlantWild

We created a wild multimodal plant disease recognition dataset PlantWild, which has the most disease categories. Our dataset introduces descriptive prompts to provide rich information in the text modality.

Source
github
Created
Jul 17, 2024
Updated
Aug 7, 2024
Signals
582 views
Availability
Linked source ready
Overview

Dataset description and usage context

PlantWild Dataset Overview

Introduction

PlantWild is a wild multimodal plant disease recognition dataset with the largest number of disease categories. The dataset introduces descriptive prompts to provide rich textual information.

Access

The dataset can be accessed via the following link: PlantWild.

Running

Code for training and evaluation is provided in main.py.

python main.py --config <CONFIG_DIR>

Results

Performance results are shown below:

Citation

If the dataset is useful to your work, please cite:

@inproceedings{MVPDR,
  title={Benchmarking In-the-Wild Multimodal Plant Disease Recognition and A Versatile Baseline},
  author={Wei, Tianqi and Chen, Zhi and Huang, Zi and Yu, Xin},
  booktitle={ACM International Conference on Multimedia},
  year={2024}
}
Need downstream help?

Pair the dataset with AI analysis and content workflows.

Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.

Explore AI studio