CyberHarem/sado_kantaicollection

---
license: mit
task_categories:
- text-to-image
tags:
- art
- not-for-all-audiences
size_categories:
- n<1K
---

# Dataset of sado (Kantai Collection)

This is the dataset of sado (Kantai Collection), containing 257 images and their tags.

The core tags of this character are `blue_hair, multicolored_hair, gradient_hair, blonde_hair, long_hair, red_eyes, hat, white_headwear, sailor_hat, fang, ahoge`, which are pruned in this dataset.

Images are crawled from many sites (e.g. danbooru, pixiv, zerochan ...), the auto-crawling system is powered by [DeepGHS Team](https://github.com/deepghs)([huggingface organization](https://huggingface.co/deepghs)).

## List of Packages

| Name | Images | Size | Download | Type | Description |
|:-----------------|---------:|:-----------|:-----------------------------------------------------------------------------------------------------------------------|:-----------|:---------------------------------------------------------------------|
| raw | 257 | 166.78 MiB | [Download](https://huggingface.co/datasets/CyberHarem/sado_kantaicollection/resolve/main/dataset-raw.zip) | Waifuc-Raw | Raw data with meta information (min edge aligned to 1400 if larger). |
| 800 | 257 | 121.61 MiB | [Download](https://huggingface.co/datasets/CyberHarem/sado_kantaicollection/resolve/main/dataset-800.zip) | IMG+TXT | dataset with the shorter side not exceeding 800 pixels. |
| stage3-p480-800 | 489 | 236.94 MiB | [Download](https://huggingface.co/datasets/CyberHarem/sado_kantaicollection/resolve/main/dataset-stage3-p480-800.zip) | IMG+TXT | 3-stage cropped dataset with the area not less than 480x480 pixels. |
| 1200 | 257 | 158.28 MiB | [Download](https://huggingface.co/datasets/CyberHarem/sado_kantaicollection/resolve/main/dataset-1200.zip) | IMG+TXT | dataset with the shorter side not exceeding 1200 pixels. |
| stage3-p480-1200 | 489 | 299.10 MiB | [Download](https://huggingface.co/datasets/CyberHarem/sado_kantaicollection/resolve/main/dataset-stage3-p480-1200.zip) | IMG+TXT | 3-stage cropped dataset with the area not less than 480x480 pixels. |

### Load Raw Dataset with Waifuc

We provide raw dataset (including tagged images) for [waifuc](https://deepghs.github.io/waifuc/main/tutorials/installation/index.html) loading. If you need this, just run the following code

```python
import os
import zipfile

from huggingface_hub import hf_hub_download
from waifuc.source import LocalSource

# download raw archive file
zip_file = hf_hub_download(
repo_id='CyberHarem/sado_kantaicollection',
repo_type='dataset',
filename='dataset-raw.zip',
)

# extract files to your directory
dataset_dir = 'dataset_dir'
os.makedirs(dataset_dir, exist_ok=True)
with zipfile.ZipFile(zip_file, 'r') as zf:
zf.extractall(dataset_dir)

# load the dataset with waifuc
source = LocalSource(dataset_dir)
for item in source:
print(item.image, item.meta['filename'], item.meta['tags'])
```

## List of Clusters

List of tag clustering result, maybe some outfits can be mined here.

### Raw Text Version

| # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | Tags |
|----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 0 | 11 |  |  |  |  |  | 1girl, blue_neckerchief, serafuku, simple_background, sleeveless_shirt, solo, undershirt, white_gloves, blue_sailor_collar, upper_body, white_background, grin, looking_at_viewer, blue_ribbon |
| 1 | 6 |  |  |  |  |  | 1girl, blue_neckerchief, blue_ribbon, blue_sailor_collar, blue_skirt, one-hour_drawing_challenge, pleated_skirt, serafuku, simple_background, sleeveless_shirt, solo, twitter_username, undershirt, white_background, white_gloves, black_pantyhose, cowboy_shot, looking_at_viewer, open_mouth |
| 2 | 5 |  |  |  |  |  | 1girl, black_pantyhose, blue_neckerchief, blue_sailor_collar, blue_skirt, open_mouth, pleated_skirt, serafuku, sleeveless_shirt, solo, undershirt, white_gloves, blue_ribbon, full_body, closed_eyes, running, sparkle |
| 3 | 5 |  |  |  |  |  | 1girl, blush, looking_at_viewer, open_mouth, solo, ass, black_pantyhose, aqua_hair, bangs, hair_between_eyes, panties_under_pantyhose, simple_background, undershirt, white_gloves, cowboy_shot, hair_down, navel, skin_fang, smile, white_background, white_panties |
| 4 | 8 |  |  |  |  |  | 1girl, solo, black_pantyhose, hood_up, raincoat, simple_background, blush, looking_at_viewer, open_mouth, smile, white_background, white_gloves, boots, full_body, long_sleeves, skin_fang, standing, bangs |
| 5 | 10 |  |  |  |  |  | 1girl, solo, black_pantyhose, long_sleeves, open_mouth, alternate_costume, christmas, smile, fur_trim, red_dress, red_footwear, santa_costume, bell, blush, looking_at_viewer, red_mittens, sack, boots, full_body, red_gloves, skin_fang, hair_between_eyes, holding, sailor_collar, simple_background |
| 6 | 5 |  |  |  |  |  | 1girl, alternate_costume, open_mouth, skin_fang, solo, blush, full_body, simple_background, white_background, long_sleeves, ritual_baton, smile, tate_eboshi, wide_sleeves, hair_between_eyes, holding, kimono, looking_at_viewer, short_hair, socks, tabi |
| 7 | 6 |  |  |  |  |  | 1girl, solo, flat_chest, grin, alternate_costume, artist_logo, blush, cowboy_shot, dated, looking_at_viewer, simple_background, white_background, collarbone |

### Table Version

| # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | 1girl | blue_neckerchief | serafuku | simple_background | sleeveless_shirt | solo | undershirt | white_gloves | blue_sailor_collar | upper_body | white_background | grin | looking_at_viewer | blue_ribbon | blue_skirt | one-hour_drawing_challenge | pleated_skirt | twitter_username | black_pantyhose | cowboy_shot | open_mouth | full_body | closed_eyes | running | sparkle | blush | ass | aqua_hair | bangs | hair_between_eyes | panties_under_pantyhose | hair_down | navel | skin_fang | smile | white_panties | hood_up | raincoat | boots | long_sleeves | standing | alternate_costume | christmas | fur_trim | red_dress | red_footwear | santa_costume | bell | red_mittens | sack | red_gloves | holding | sailor_collar | ritual_baton | tate_eboshi | wide_sleeves | kimono | short_hair | socks | tabi | flat_chest | artist_logo | dated | collarbone |
|----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------|:-------------------|:-----------|:--------------------|:-------------------|:-------|:-------------|:---------------|:---------------------|:-------------|:-------------------|:-------|:--------------------|:--------------|:-------------|:-----------------------------|:----------------|:-------------------|:------------------|:--------------|:-------------|:------------|:--------------|:----------|:----------|:--------|:------|:------------|:--------|:--------------------|:--------------------------|:------------|:--------|:------------|:--------|:----------------|:----------|:-----------|:--------|:---------------|:-----------|:--------------------|:------------|:-----------|:------------|:---------------|:----------------|:-------|:--------------|:-------|:-------------|:----------|:----------------|:---------------|:--------------|:---------------|:---------|:-------------|:--------|:-------|:-------------|:--------------|:--------|:-------------|
| 0 | 11 |  |  |  |  |  | X | X | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 1 | 6 |  |  |  |  |  | X | X | X | X | X | X | X | X | X | | X | | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 2 | 5 |  |  |  |  |  | X | X | X | | X | X | X | X | X | | | | | X | X | | X | | X | | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 3 | 5 |  |  |  |  |  | X | | | X | | X | X | X | | | X | | X | | | | | | X | X | X | | | | | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 4 | 8 |  |  |  |  |  | X | | | X | | X | | X | | | X | | X | | | | | | X | | X | X | | | | X | | | X | | | | | X | X | | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | |
| 5 | 10 |  |  |  |  |  | X | | | X | | X | | | | | | | X | | | | | | X | | X | X | | | | X | | | | X | | | | X | X | | | | X | X | | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | |
| 6 | 5 |  |  |  |  |  | X | | | X | | X | | | | | X | | X | | | | | | | | X | X | | | | X | | | | X | | | | X | X | | | | | X | | X | | | | | | | | | | X | | X | X | X | X | X | X | X | | | | |
| 7 | 6 |  |  |  |  |  | X | | | X | | X | | | | | X | X | X | | | | | | | X | | | | | | X | | | | | | | | | | | | | | | | X | | | | | | | | | | | | | | | | | | | X | X | X | X |

原始信息汇总

数据集概述

数据集名称

Dataset of sado (Kantai Collection)

数据集描述

该数据集包含257张图片及其标签，主要围绕角色sado（舰队收藏）。图片来源于多个网站，如danbooru、pixiv、zerochan等。

核心标签

• blue_hair
• multicolored_hair
• gradient_hair
• blonde_hair
• long_hair
• red_eyes
• hat
• white_headwear
• sailor_hat
• fang
• ahoge

数据集包列表

名称	图片数量	大小	类型	描述
raw	257	166.78 MiB	Waifuc-Raw	包含元信息的原始数据（最小边对齐到1400像素，如果更大）
800	257	121.61 MiB	IMG+TXT	短边不超过800像素的数据集
stage3-p480-800	489	236.94 MiB	IMG+TXT	3阶段裁剪数据集，区域不小于480x480像素
1200	257	158.28 MiB	IMG+TXT	短边不超过1200像素的数据集
stage3-p480-1200	489	299.10 MiB	IMG+TXT	3阶段裁剪数据集，区域不小于480x480像素

标签聚类结果

原始文本版本

#	样本数量	图片示例1	图片示例2	图片示例3	图片示例4	图片示例5	标签
0	11						1girl, blue_neckerchief, serafuku, simple_background, sleeveless_shirt, solo, undershirt, white_gloves, blue_sailor_collar, upper_body, white_background, grin, looking_at_viewer, blue_ribbon
1	6						1girl, blue_neckerchief, blue_ribbon, blue_sailor_collar, blue_skirt, one-hour_drawing_challenge, pleated_skirt, serafuku, simple_background, sleeveless_shirt, solo, twitter_username, undershirt, white_background, white_gloves, black_pantyhose, cowboy_shot, looking_at_viewer, open_mouth
2	5						1girl, black_pantyhose, blue_neckerchief, blue_sailor_collar, blue_skirt, open_mouth, pleated_skirt, serafuku, sleeveless_shirt, solo, undershirt, white_gloves, blue_ribbon, full_body, closed_eyes, running, sparkle
3	5						1girl, blush, looking_at_viewer, open_mouth, solo, ass, black_pantyhose, aqua_hair, bangs, hair_between_eyes, panties_under_pantyhose, simple_background, undershirt, white_gloves, cowboy_shot, hair_down, navel, skin_fang, smile, white_background, white_panties
4	8						1girl, solo, black_pantyhose, hood_up, raincoat, simple_background, blush, looking_at_viewer, open_mouth, smile, white_background, white_gloves, boots, full_body, long_sleeves, skin_fang, standing, bangs
5	10						1girl, solo, black_pantyhose, long_sleeves, open_mouth, alternate_costume, christmas, smile, fur_trim, red_dress, red_footwear, santa_costume, bell, blush, looking_at_viewer, red_mittens, sack, boots, full_body, red_gloves, skin_fang, hair_between_eyes, holding, sailor_collar, simple_background
6	5						1girl, alternate_costume, open_mouth, skin_fang, solo, blush, full_body, simple_background, white_background, long_sleeves, ritual_baton, smile, tate_eboshi, wide_sleeves, hair_between_eyes, holding, kimono, looking_at_viewer, short_hair, socks, tabi
7	6						1girl, solo, flat_chest, grin, alternate_costume, artist_logo, blush, cowboy_shot, dated, looking_at_viewer, simple_background, white_background, collarbone

表格版本

#	样本数量	图片示例1	图片示例2	图片示例3	图片示例4	图片示例5	1girl	blue_neckerchief	serafuku	simple_background	sleeveless_shirt	solo	undershirt	white_gloves	blue_sailor_collar	upper_body	white_background	grin	looking_at_viewer	blue_ribbon	blue_skirt	one-hour_drawing_challenge	pleated_skirt	twitter_username	black_pantyhose	cowboy_shot	open_mouth	full_body	closed_eyes	running	sparkle	blush	ass	aqua_hair	bangs	hair_between_eyes	panties_under_pantyhose	hair_down	navel	skin_fang	smile	white_panties	hood_up	raincoat	boots	long_sleeves	standing	alternate_costume	christmas	fur_trim	red_dress	red_footwear	santa_costume	bell	red_mittens	sack	red_gloves	holding	sailor_collar	ritual_baton	tate_eboshi	wide_sleeves	kimono	short_hair	socks	tabi	flat_chest	artist_logo	dated	collarbone
0	11						X	X	X	X	X	X	X	X	X	X	X	X	X	X
1	6						X	X	X	X	X	X	X	X	X		X		X	X	X	X	X	X	X	X	X
2	5						X	X	X		X	X	X	X	X					X	X		X		X		X	X	X	X	X
3	5						X			X		X	X	X			X		X						X	X	X					X	X	X	X	X	X	X	X	X	X	X
4	8						X			X		X		X			X		X						X		X	X				X			X					X	X		X	X	X	X	X
5	10						X			X		X							X						X		X	X				X				X				X	X				X	X		X	X	X	X	X	X	X	X	X	X	X	X
6	5	![](samples/6/clu6-sample0.png

搜集汇总

数据集介绍

构建方式

在动漫角色数据集构建领域，CyberHarem/sado_kantaicollection 专注于收录《舰队Collection》中的角色 sado。该数据集通过自动化爬取系统从 Danbooru、Pixiv、Zerochan 等多个知名图像站点采集图像，爬取系统由 DeepGHS 团队提供技术支持。原始数据包含 257 张图像及其对应的标签信息，核心标签涵盖蓝发、渐变色发、红瞳、帽子、虎牙等关键特征，并在数据集中进行了精简处理。数据集提供多种处理版本，包括原始数据包（含元信息，最小边对齐至 1400 像素）以及短边不超过 800 或 1200 像素的标准化版本，同时还有经过三阶段裁剪且面积不低于 480×480 像素的增强版本。

使用方法

数据集的使用方式灵活多样。对于原始数据，推荐通过 waifuc 库进行加载，用户可先使用 huggingface_hub 下载压缩包，解压后利用 LocalSource 接口逐项读取图像及其元数据与标签。对于标准化版本，可直接解压后作为文本到图像模型的训练输入，图像与对应标签以 IMG+TXT 格式存储。研究人员还可参考提供的聚类结果，选择特定簇的数据进行针对性训练或分析。此外，数据集采用 MIT 许可证发布，允许自由使用和修改，但需注意内容可能不适用于所有观众。

背景与挑战

背景概述

在文本到图像生成领域，高质量、标注精细的角色数据集对于训练能够忠实再现特定角色特征的模型至关重要。CyberHarem/sado_kantaicollection 数据集由 DeepGHS 团队于近年创建，专注于《舰队Collection》中的角色“sado”。该数据集包含 257 张图像及其对应的标签，核心标签涵盖蓝发、多色发、渐变发、红瞳、帽子、虎牙等关键特征，为角色定制化生成提供了精准的语义锚点。图像来源涵盖 Danbooru、Pixiv、Zerochan 等多个平台，借助自动化爬取系统构建而成。该数据集的影响力在于，它为二次元角色生成任务树立了从数据采集到标签精简的标准化流程，推动了特定角色在生成模型中的可控性与保真度研究。

当前挑战

该数据集所面临的挑战主要体现在两个层面。在领域问题层面，文本到图像生成任务中，角色特征的精确还原始终是核心难点，尤其是面对“sado”这样具有复杂发型、多色渐变发色及多种服饰变体的角色时，模型需在保持身份一致性的同时，应对不同姿态、场景和服装组合带来的歧义。在构建过程层面，数据采集面临跨平台图像质量参差不齐、标签噪声与冗余、以及版权合规性等问题；此外，数据量仅 257 张，规模较小，可能限制模型泛化能力，而标签的裁剪与聚类虽提升了可用性，但可能丢失部分细粒度信息，影响对稀有属性的学习效果。

常用场景

经典使用场景

在二次元角色定制与生成领域，CyberHarem/sado_kantaicollection数据集凭借其包含的257张高分辨率图像及精细标注标签，成为训练文本到图像模型（如Stable Diffusion）的核心素材。研究者常利用该数据集进行角色概念的微调（DreamBooth/LoRA），通过其提供的多尺度裁剪版本（如800px与1200px）和聚类标签，实现对特定角色（如舰娘“佐渡”）的风格化再现与属性解耦。

解决学术问题

该数据集有效解决了动漫角色生成中身份一致性保持的难题，为少样本学习场景下的概念注入提供了标准化基准。通过其标签聚类结果，研究者可系统分析不同服饰、姿态与背景组合对生成效果的影响，推动了对多模态特征对齐与语义可控生成机制的深入理解，显著提升了模型在细粒度角色属性上的泛化能力。

实际应用

在实际应用中，该数据集被广泛用于游戏角色概念设计、虚拟主播形象定制以及同人创作辅助工具的开发。创作者可通过加载预处理后的数据集，快速生成符合角色核心特征（如蓝发、虎牙、水手帽）的多样化图像，并利用其裁剪版本适配不同分辨率需求，从而降低手工绘制成本，加速创意迭代流程。

数据集最近研究

最新研究方向

在二次元角色图像生成领域，针对特定角色的精细化数据集构建正成为推动文本到图像模型发展的关键环节。该数据集聚焦于《舰队Collection》中的角色“sado”，通过整合来自Danbooru、Pixiv等多平台的高质量图像资源，并辅以自动化的标签标注与聚类分析，为角色定制化生成提供了标准化训练素材。前沿研究方向集中于利用此类细粒度数据集优化扩散模型的角色一致性表现，尤其是在处理复杂发色渐变、服饰细节及多姿态渲染等挑战时展现出显著价值。同时，数据集中包含的多尺度裁剪版本与原始元数据，为研究图像质量与模型泛化能力之间的平衡提供了实证基础，其聚类结果亦有助于挖掘角色在不同场景下的视觉特征规律，对推动动漫风格生成模型的精细化发展具有重要参考意义。