Datasets:
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc0-1.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- la
|
| 6 |
+
source_datasets:
|
| 7 |
+
- imageomics/TreeOfLife-10M
|
| 8 |
+
task_categories:
|
| 9 |
+
- image-classification
|
| 10 |
+
- zero-shot-classification
|
| 11 |
+
pretty_name: TreeOfLife-10M WEBP
|
| 12 |
+
size_categories:
|
| 13 |
+
- 10M<n<100M
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
# Dataset Card for TreeOfLife-10M-WEBP
|
| 17 |
+
|
| 18 |
+
## Dataset Description
|
| 19 |
+
|
| 20 |
+
This is an optimized version of the [TreeOfLife-10M](https://huggingface.co/datasets/imageomics/TreeOfLife-10M) dataset,
|
| 21 |
+
containing over 10 million images covering 454 thousand taxa in the tree of life.
|
| 22 |
+
This version has been processed to improve usability and reduce storage requirements while maintaining full compatibility with the original dataset structure.
|
| 23 |
+
|
| 24 |
+
### Dataset Summary
|
| 25 |
+
|
| 26 |
+
This version modifies the original dataset as follows:
|
| 27 |
+
|
| 28 |
+
- Corrupted files were repaired.
|
| 29 |
+
- Very large images (some >40K pixels in width) were resized so that the total number of pixels < 1,048,576 (=1024×1024), preserving aspect ratio.
|
| 30 |
+
- All images were re-encoded to WEBP format.
|
| 31 |
+
- The dataset was repacked in the same shard structure as the original to remain fully compatible.
|
| 32 |
+
|
| 33 |
+
The result is a significantly reduced dataset size (~500GB vs. ~2TB), with lower I/O overhead and fewer extreme image cases that can slow down training pipelines.
|
| 34 |
+
|
| 35 |
+
For dataset details, licensing information, taxonomy information and annotation process, please see the [original dataset card](https://huggingface.co/datasets/imageomics/TreeOfLife-10M).
|
| 36 |
+
|
| 37 |
+
## Limitations
|
| 38 |
+
|
| 39 |
+
- Maintains all original dataset limitations regarding taxonomic coverage and class imbalance
|
| 40 |
+
- Some images have been resized, which may affect fine-grained visual analysis of extremely high-resolution specimens
|
| 41 |
+
|
| 42 |
+
## Licensing
|
| 43 |
+
|
| 44 |
+
This repackaged dataset is distributed under the same licensing terms as the original TreeOfLife-10M dataset.
|
| 45 |
+
Please review the [original licensing information](https://huggingface.co/datasets/imageomics/TreeOfLife-10M#licensing-information) before using this dataset, as all terms and restrictions remain applicable.
|
| 46 |
+
|
| 47 |
+
## Citation
|
| 48 |
+
|
| 49 |
+
```bibtex
|
| 50 |
+
@dataset{treeoflife_10m,
|
| 51 |
+
author = {Samuel Stevens and Jiaman Wu and Matthew J Thompson and Elizabeth G Campolongo and Chan Hee Song and David Edward Carlyn and Li Dong and Wasila M Dahdul and Charles Stewart and Tanya Berger-Wolf and Wei-Lun Chao and Yu Su},
|
| 52 |
+
title = {TreeOfLife-10M},
|
| 53 |
+
year = {2023},
|
| 54 |
+
url = {https://huggingface.co/datasets/imageomics/TreeOfLife-10M},
|
| 55 |
+
doi = {10.57967/hf/1972},
|
| 56 |
+
publisher = {Hugging Face}
|
| 57 |
+
}
|
| 58 |
+
|
| 59 |
+
@inproceedings{stevens2024bioclip,
|
| 60 |
+
title = {{B}io{CLIP}: A Vision Foundation Model for the Tree of Life},
|
| 61 |
+
author = {Samuel Stevens and Jiaman Wu and Matthew J Thompson and Elizabeth G Campolongo and Chan Hee Song and David Edward Carlyn and Li Dong and Wasila M Dahdul and Charles Stewart and Tanya Berger-Wolf and Wei-Lun Chao and Yu Su},
|
| 62 |
+
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
|
| 63 |
+
year = {2024},
|
| 64 |
+
pages = {19412-19424}
|
| 65 |
+
}
|
| 66 |
+
```
|
| 67 |
+
|
| 68 |
+
## Acknowledgments
|
| 69 |
+
|
| 70 |
+
This optimization builds upon the outstanding work of the original TreeOfLife-10M creators.
|
| 71 |
+
All credit for data curation, taxonomic labeling, and scientific contributions belongs to the original team at the Imageomics Institute.
|