p1atdev/FractalDB-60

---
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': a1
'1': a2
'2': a3
'3': a4
'4': ammonite
'5': bamboofern
'6': bedder
'7': binary
'8': branch
'9': broccoli
'10': bud
'11': c_curve
'12': castle
'13': cedarleaf
'14': coral
'15': crystal
'16': deerfern
'17': dragon_curve
'18': drumlin
'19': fern
'20': filmyfern
'21': fleabane
'22': flower
'23': gaku
'24': ginkgo
'25': gold_dragon
'26': grassfern
'27': greygoldenrod
'28': groundpine
'29': involucre
'30': koch_curve
'31': koch_snowflake
'32': maple_leaf
'33': mcWorter_pedigree
'34': morningglory
'35': newyorkfern
'36': octopuslegs
'37': penta
'38': pinetree
'39': rose
'40': shieldfern
'41': sierpinski_carpet
'42': sierpinski_gasket
'43': sierpinski_pentagon
'44': snail
'45': snowcap
'46': snowdrift
'47': spiderbrake
'48': spiral
'49': spleenwort_fern
'50': star
'51': sticks
'52': sunflower
'53': supernova
'54': swirl
'55': tree
'56': turbanshell
'57': umbrellafern
'58': watersprite
'59': zigzag
splits:
- name: train
num_bytes: 3588623140
num_examples: 60000
download_size: 1829671228
dataset_size: 3588623140
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: cc-by-4.0
task_categories:
- image-classification
pretty_name: 'FractalDB 60 '
size_categories:
- 10K<n<100K
---

# FractalDB 60

FractalDB 60 dataset from [Pre-training without Natural Images](https://hirokatsukataoka16.github.io/Pretraining-without-Natural-Images/).

[Original repo](https://github.com/hirokatsukataoka16/FractalDB-Pretrained-ResNet-PyTorch) | [Project page](https://hirokatsukataoka16.github.io/Pretraining-without-Natural-Images/) | [arXiv](https://arxiv.org/abs/2101.08515)

## Citing

```bibtex
@article{KataokaIJCV2022,
author={Kataoka, Hirokatsu and Okayasu, Kazushige and Matsumoto, Asato and Yamagata, Eisuke and Yamada, Ryosuke and Inoue, Nakamasa and Nakamura, Akio and Satoh, Yutaka},
title={Pre-training without Natural Images},
article={International Journal on Computer Vision (IJCV)},
year={2022},
}

@inproceedings{KataokaACCV2020,
author={Kataoka, Hirokatsu and Okayasu, Kazushige and Matsumoto, Asato and Yamagata, Eisuke and Yamada, Ryosuke and Inoue, Nakamasa and Nakamura, Akio and Satoh, Yutaka},
title={Pre-training without Natural Images},
booktitle={Asian Conference on Computer Vision (ACCV)},
year={2020},
}

@misc{kataoka2021pretraining,
title={Pre-training without Natural Images},
author={Hirokatsu Kataoka and Kazushige Okayasu and Asato Matsumoto and Eisuke Yamagata and Ryosuke Yamada and Nakamasa Inoue and Akio Nakamura and Yutaka Satoh},
year={2021},
eprint={2101.08515},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```

原始信息汇总

FractalDB 60 数据集概述

数据集信息

特征

• 图像 (image): 数据类型为图像。
• 标签 (label): 数据类型为类别标签，包含以下类别名称：
  • 0: a1
  • 1: a2
  • 2: a3
  • 3: a4
  • 4: ammonite
  • 5: bamboofern
  • 6: bedder
  • 7: binary
  • 8: branch
  • 9: broccoli
  • 10: bud
  • 11: c_curve
  • 12: castle
  • 13: cedarleaf
  • 14: coral
  • 15: crystal
  • 16: deerfern
  • 17: dragon_curve
  • 18: drumlin
  • 19: fern
  • 20: filmyfern
  • 21: fleabane
  • 22: flower
  • 23: gaku
  • 24: ginkgo
  • 25: gold_dragon
  • 26: grassfern
  • 27: greygoldenrod
  • 28: groundpine
  • 29: involucre
  • 30: koch_curve
  • 31: koch_snowflake
  • 32: maple_leaf
  • 33: mcWorter_pedigree
  • 34: morningglory
  • 35: newyorkfern
  • 36: octopuslegs
  • 37: penta
  • 38: pinetree
  • 39: rose
  • 40: shieldfern
  • 41: sierpinski_carpet
  • 42: sierpinski_gasket
  • 43: sierpinski_pentagon
  • 44: snail
  • 45: snowcap
  • 46: snowdrift
  • 47: spiderbrake
  • 48: spiral
  • 49: spleenwort_fern
  • 50: star
  • 51: sticks
  • 52: sunflower
  • 53: supernova
  • 54: swirl
  • 55: tree
  • 56: turbanshell
  • 57: umbrellafern
  • 58: watersprite
  • 59: zigzag

数据分割

• 训练集 (train): 包含60000个样本，占用3588623140字节。

数据集大小

• 下载大小: 1829671228字节
• 数据集大小: 3588623140字节

配置

• 默认配置 (default):
  • 数据文件路径: data/train-*

许可

• 许可证: CC BY 4.0

任务类别

• 图像分类

数据集名称

• FractalDB 60

数据集规模

• 10K<n<100K

搜集汇总

背景与挑战

背景概述

FractalDB-60是一个用于图像分类任务的数据集，包含60,000张512x512像素的图像，覆盖60个类别，总大小1.83 GB。该数据集基于论文'Pre-training without Natural Images'（arXiv:2101.08515），专为无需自然图像的模型预训练设计，采用CC BY 4.0许可证。