# Classification of plants of Southeast Asia — Data

来源：https://www.kaggle.com/competitions/classification-of-plants-of-southeast-asia/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

Bali26 (https://realtechsupport.org/projects/bali26_overview_bahasa.html) is an image dataset dedicated to ethnobotany, the study of the interaction between people and plants. Bali26 is the first machine-vision ready image collection of ethnobotanically significant flora of south-east Asia, collected on the island of Bali in 2020 (and amended in 2021) together with Balinese residents with intimate local knowledge in coordination with expertise from the Indonesia National Research and Innovation Agency.

The images were created from video documentation of flora in the wild with Catch & Release (https://github.com/realtechsupport/c-plus-r). The images show a deliriously rich flora with more information content (entropy) than similar examples in ImageNet. Not unlike the IRIS dataset, the data samples of Bali26 were collected in a controlled real-world environment; in the wild, in a restricted area, within a short time frame, by a small group of data collectors. Bali26 contains 26 categories with approximately 2000 images per category. 

The publications listed below describe the curation of this dataset and some early classification experiments.

·         M. Böhlen and W. Sujarwo, "Return to Bali," 2020 Second International Conference on Transdisciplinary AI (TransAI), 2020, pp. 92-95.

·         M. Böhlen and W. Sujarwo, "Machine Learning in Ethnobotany," 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2020, pp. 108-113.

·         M. Böhlen, R. Jain, W. Sujarwo and V. Chandola, "From images in the wild to video-informed image classification," 2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA), 2021, pp. 656-661.


## Files

*   **bali-26_train** - the training set, which includes 26 folders, each contains images of the same category
*   **bali-26_test** - the test set, which includes test images without folder structures
*   **sample_submission.csv** - a sample submission file in the correct format