CyberHarem/ch_en_arknights

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

# Dataset of ch'en/チェン/陈 (Arknights)

This is the dataset of ch'en/チェン/陈 (Arknights), containing 500 images and their tags.

The core tags of this character are `horns, blue_hair, dragon_horns, long_hair, red_eyes, hair_between_eyes, breasts, dragon_tail, tail, medium_breasts, twintails, sidelocks`, 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 | 500 | 1008.34 MiB | [Download](https://huggingface.co/datasets/CyberHarem/ch_en_arknights/resolve/main/dataset-raw.zip) | Waifuc-Raw | Raw data with meta information (min edge aligned to 1400 if larger). |
| 800 | 500 | 449.18 MiB | [Download](https://huggingface.co/datasets/CyberHarem/ch_en_arknights/resolve/main/dataset-800.zip) | IMG+TXT | dataset with the shorter side not exceeding 800 pixels. |
| stage3-p480-800 | 1337 | 1.01 GiB | [Download](https://huggingface.co/datasets/CyberHarem/ch_en_arknights/resolve/main/dataset-stage3-p480-800.zip) | IMG+TXT | 3-stage cropped dataset with the area not less than 480x480 pixels. |
| 1200 | 500 | 824.33 MiB | [Download](https://huggingface.co/datasets/CyberHarem/ch_en_arknights/resolve/main/dataset-1200.zip) | IMG+TXT | dataset with the shorter side not exceeding 1200 pixels. |
| stage3-p480-1200 | 1337 | 1.58 GiB | [Download](https://huggingface.co/datasets/CyberHarem/ch_en_arknights/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/ch_en_arknights',
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 | 26 |  |  |  |  |  | 1girl, holding_sword, short_shorts, solo, white_shirt, black_gloves, black_jacket, black_shorts, looking_at_viewer, open_jacket, yellow_necktie, fingerless_gloves, sleeveless_shirt, bare_shoulders, off_shoulder, cowboy_shot, belt, thighs, navel, collared_shirt, standing, low_twintails |
| 1 | 5 |  |  |  |  |  | 1girl, black_footwear, black_gloves, black_jacket, black_shorts, fingerless_gloves, looking_at_viewer, open_jacket, shin_guards, short_shorts, solo, white_shirt, yellow_necktie, boots, closed_mouth, full_body, holding_sword, low_twintails, collared_shirt, thighs, sitting, standing |
| 2 | 22 |  |  |  |  |  | 1girl, solo, white_shirt, yellow_necktie, upper_body, black_jacket, looking_at_viewer, open_jacket, off_shoulder, sleeveless_shirt, bare_shoulders, collared_shirt, low_twintails, closed_mouth, simple_background, white_background |
| 3 | 12 |  |  |  |  |  | 1girl, bare_shoulders, black_bikini, blue_sky, cloud, day, navel, official_alternate_costume, outdoors, solo, standing, stomach, bare_arms, baseball_cap, cowboy_shot, horns_through_headwear, looking_at_viewer, open_fly, grey_shorts, short_shorts, white_headwear, collarbone, pouch, highleg_bikini, thighs, water, halterneck, beach, belt, ocean, wristwatch |
| 4 | 8 |  |  |  |  |  | 1girl, bare_arms, bare_shoulders, black_bikini, highleg_bikini, looking_at_viewer, navel, official_alternate_costume, short_shorts, simple_background, solo, stomach, thighs, white_background, baseball_cap, horns_through_headwear, open_fly, sunglasses, cowboy_shot, grey_shorts, halterneck, holding_removed_eyewear, pouch, standing, unworn_eyewear, white_headwear, belt, folded_ponytail, heart, large_breasts, collarbone, white_shorts, wristwatch |
| 5 | 5 |  |  |  |  |  | 1girl, bare_shoulders, china_dress, cleavage_cutout, double_bun, looking_at_viewer, official_alternate_costume, sleeveless_dress, solo, red_dress, smile, upper_body, bare_arms, large_breasts, hand_up, red_background |
| 6 | 23 |  |  |  |  |  | 1girl, bare_arms, bare_shoulders, china_dress, double_bun, looking_at_viewer, official_alternate_costume, red_dress, sleeveless_dress, solo, black_shorts, thigh_strap, short_shorts, thighs, white_background, cleavage_cutout, bead_bracelet, simple_background, bare_legs, red_footwear, sitting, high_heels, hand_up, large_breasts |
| 7 | 5 |  |  |  |  |  | 1girl, black_dress, closed_mouth, enmaided, frills, large_breasts, maid_apron, simple_background, solo, white_apron, white_background, blush, dragon_girl, juliet_sleeves, looking_at_viewer, maid_headdress, cowboy_shot, from_side, holding, standing, upper_body |
| 8 | 5 |  |  |  |  |  | 1girl, closed_mouth, long_sleeves, solo, alternate_costume, looking_at_viewer, black_dress, cleavage, low_twintails, sitting, dragon_girl, playing_instrument, purple_eyes, smile, upper_body |

### Table Version

| # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | 1girl | holding_sword | short_shorts | solo | white_shirt | black_gloves | black_jacket | black_shorts | looking_at_viewer | open_jacket | yellow_necktie | fingerless_gloves | sleeveless_shirt | bare_shoulders | off_shoulder | cowboy_shot | belt | thighs | navel | collared_shirt | standing | low_twintails | black_footwear | shin_guards | boots | closed_mouth | full_body | sitting | upper_body | simple_background | white_background | black_bikini | blue_sky | cloud | day | official_alternate_costume | outdoors | stomach | bare_arms | baseball_cap | horns_through_headwear | open_fly | grey_shorts | white_headwear | collarbone | pouch | highleg_bikini | water | halterneck | beach | ocean | wristwatch | sunglasses | holding_removed_eyewear | unworn_eyewear | folded_ponytail | heart | large_breasts | white_shorts | china_dress | cleavage_cutout | double_bun | sleeveless_dress | red_dress | smile | hand_up | red_background | thigh_strap | bead_bracelet | bare_legs | red_footwear | high_heels | black_dress | enmaided | frills | maid_apron | white_apron | blush | dragon_girl | juliet_sleeves | maid_headdress | from_side | holding | long_sleeves | alternate_costume | cleavage | playing_instrument | purple_eyes |
|----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------|:----------------|:---------------|:-------|:--------------|:---------------|:---------------|:---------------|:--------------------|:--------------|:-----------------|:--------------------|:-------------------|:-----------------|:---------------|:--------------|:-------|:---------|:--------|:-----------------|:-----------|:----------------|:-----------------|:--------------|:--------|:---------------|:------------|:----------|:-------------|:--------------------|:-------------------|:---------------|:-----------|:--------|:------|:-----------------------------|:-----------|:----------|:------------|:---------------|:-------------------------|:-----------|:--------------|:-----------------|:-------------|:--------|:-----------------|:--------|:-------------|:--------|:--------|:-------------|:-------------|:--------------------------|:-----------------|:------------------|:--------|:----------------|:---------------|:--------------|:------------------|:-------------|:-------------------|:------------|:--------|:----------|:-----------------|:--------------|:----------------|:------------|:---------------|:-------------|:--------------|:-----------|:---------|:-------------|:--------------|:--------|:--------------|:-----------------|:-----------------|:------------|:----------|:---------------|:--------------------|:-----------|:---------------------|:--------------|
| 0 | 26 |  |  |  |  |  | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 1 | 5 |  |  |  |  |  | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | X | | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 2 | 22 |  |  |  |  |  | X | | | X | X | | X | | X | X | X | | X | X | X | | | | | X | | X | | | | X | | | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 3 | 12 |  |  |  |  |  | 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 | 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 | X | X | X | X | X | X | | X | | | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 5 | 5 |  |  |  |  |  | X | | | X | | | | | X | | | | | X | | | | | | | | | | | | | | | X | | | | | | | X | | | X | | | | | | | | | | | | | | | | | | | X | | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | |
| 6 | 23 |  |  |  |  |  | X | | X | X | | | | X | X | | | | | X | | | | X | | | | | | | | | | X | | X | X | | | | | X | | | X | | | | | | | | | | | | | | | | | | | X | | X | X | X | X | X | | X | | X | X | X | X | X | | | | | | | | | | | | | | | | |
| 7 | 5 |  |  |  |  |  | X | | | X | | | | | X | | | | | | | X | | | | | X | | | | | X | | | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | X | | | | | | | | | | | | | | | X | X | X | X | X | X | X | X | X | X | X | | | | | |
| 8 | 5 |  |  |  |  |  | X | | | X | | | | | X | | | | | | | | | | | | | X | | | | X | | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | X | | | | | | | | X | | | | | | X | | | | | X | X | X | X | X |

原始信息汇总

数据集概述

数据集信息

• 名称: Dataset of chen/チェン/陈 (Arknights)
• 许可证: MIT
• 任务类别: text-to-image
• 标签: art, not-for-all-audiences
• 大小类别: n<1K

数据内容

• 图像数量: 500
• 核心标签: horns, blue_hair, dragon_horns, long_hair, red_eyes, hair_between_eyes, breasts, dragon_tail, tail, medium_breasts, twintails, sidelocks

数据包列表

名称	图像数量	大小	类型	描述
raw	500	1008.34 MiB	Waifuc-Raw	原始数据，包含元信息（最小边对齐到1400像素，如果更大）。
800	500	449.18 MiB	IMG+TXT	短边不超过800像素的数据集。
stage3-p480-800	1337	1.01 GiB	IMG+TXT	3阶段裁剪数据集，区域不小于480x480像素。
1200	500	824.33 MiB	IMG+TXT	短边不超过1200像素的数据集。
stage3-p480-1200	1337	1.58 GiB	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	图像示例	标签
0	26		1girl, holding_sword, short_shorts, solo, white_shirt, black_gloves, black_jacket, black_shorts, looking_at_viewer, open_jacket, yellow_necktie, fingerless_gloves, sleeveless_shirt, bare_shoulders, off_shoulder, cowboy_shot, belt, thighs, navel, collared_shirt, standing, low_twintails
1	5		1girl, black_footwear, black_gloves, black_jacket, black_shorts, fingerless_gloves, looking_at_viewer, open_jacket, shin_guards, short_shorts, solo, white_shirt, yellow_necktie, boots, closed_mouth, full_body, holding_sword, low_twintails, collared_shirt, thighs, sitting, standing
2	22		1girl, solo, white_shirt, yellow_necktie, upper_body, black_jacket, looking_at_viewer, open_jacket, off_shoulder, sleeveless_shirt, bare_shoulders, collared_shirt, low_twintails, closed_mouth, simple_background, white_background
3	12		1girl, bare_shoulders, black_bikini, blue_sky, cloud, day, navel, official_alternate_costume, outdoors, solo, standing, stomach, bare_arms, baseball_cap, cowboy_shot, horns_through_headwear, looking_at_viewer, open_fly, grey_shorts, short_shorts, white_headwear, collarbone, pouch, highleg_bikini, thighs, water, halterneck, beach, belt, ocean, wristwatch
4	8		1girl, bare_arms, bare_shoulders, black_bikini, highleg_bikini, looking_at_viewer, navel, official_alternate_costume, short_shorts, simple_background, solo, stomach, thighs, white_background, baseball_cap, horns_through_headwear, open_fly, sunglasses, cowboy_shot, grey_shorts, halterneck, holding_removed_eyewear, pouch, standing, unworn_eyewear, white_headwear, belt, folded_ponytail, heart, large_breasts, collarbone, white_shorts, wristwatch
5	5		1girl, bare_shoulders, china_dress, cleavage_cutout, double_bun, looking_at_viewer, official_alternate_costume, sleeveless_dress, solo, red_dress, smile, upper_body, bare_arms, large_breasts, hand_up, red_background
6	23		1girl, bare_arms, bare_shoulders, china_dress, double_bun, looking_at_viewer, official_alternate_costume, red_dress, sleeveless_dress, solo, black_shorts, thigh_strap, short_shorts, thighs, white_background, cleavage_cutout, bead_bracelet, simple_background, bare_legs, red_footwear, sitting, high_heels, hand_up, large_breasts
7	5		1girl, black_dress, closed_mouth, enmaided, frills, large_breasts, maid_apron, simple_background, solo, white_apron, white_background, blush, dragon_girl, juliet_sleeves, looking_at_viewer, maid_headdress, cowboy_shot, from_side, holding, standing, upper_body
8	5		1girl, closed_mouth, long_sleeves, solo, alternate_costume, looking_at_viewer, black_dress, cleavage, low_twintails, sitting, dragon_girl, playing_instrument, purple_eyes, smile, upper_body

搜集汇总

数据集介绍

构建方式

在二次元角色数据集构建领域，针对《明日方舟》中'陈'这一角色，CyberHarem团队精心打造了包含500幅图像及其对应标签的专业数据集。该数据集通过自动化爬取系统从Danbooru、Pixiv、Zerochan等多个知名插画平台采集原始图像，系统由DeepGHS团队研发的Waifuc框架驱动。数据集剔除了角色核心标签（如龙角、蓝发、长尾等），保留了多样化的服饰与场景特征。数据包提供多种预处理版本：原始数据（raw）保留元信息且最小边对齐至1400像素；800与1200版本分别限制短边不超过对应像素；stage3系列则采用三阶段裁剪策略，确保裁剪区域不低于480×480像素，有效提升图像质量与训练效率。

使用方法

数据集的调用方式灵活多样，适配不同研究需求。用户可通过HuggingFace Hub直接下载预处理后的ZIP压缩包，解压后即可获得图像与对应TXT标签文件。对于需要原始元数据的场景，推荐使用Waifuc框架加载raw版本：首先通过huggingface_hub库下载dataset-raw.zip，解压至本地目录，随后利用LocalSource类迭代访问每个数据项，轻松获取图像、文件名及标签信息。此外，数据集提供的聚类表格可直接作为先验知识用于少样本学习或风格迁移任务，研究者可根据聚类ID快速定位特定服饰组合的图像子集。

背景与挑战

背景概述

在文本到图像生成领域，高质量、细粒度的角色数据集对于模型学习特定角色的视觉特征与语义标签之间的映射关系至关重要。CyberHarem/ch_en_arknights 数据集由 DeepGHS 团队于近期创建，聚焦于游戏《明日方舟》中的人气角色“陈”，收录了 500 张经过多源爬取与清洗的图像及其对应的标签信息。该数据集的核心研究问题在于如何通过自动化的爬取与标注流程，构建一个覆盖角色多姿态、多服装变体（如默认装束、泳装、旗袍、女仆装等）的标准化图像-文本对集合，从而支撑扩散模型等生成式架构的微调与评估。其影响力主要体现在为二次元角色生成任务提供了可复现的基准数据，并推动了角色一致性生成方向的技术探索。

当前挑战

当前数据集面临的核心挑战包括：其一，在领域问题层面，尽管数据集覆盖了角色的多种外观，但仅包含单一角色且样本量较小（500 张），难以应对生成任务中对于角色多角度、复杂场景与光照变化下的泛化需求，易导致模型过拟合或生成结果缺乏多样性。其二，在构建过程中，图像来源于 Danbooru、Pixiv 等社区站点，自动爬取系统虽提高了效率，但引入了标签噪声与不一致性（如不同来源对同一属性的标注差异），且缺乏对图像版权与质量的有效筛选机制。此外，多阶段裁剪与缩放流程（如 stage3-p480-800）虽适配不同分辨率需求，但可能因裁剪策略导致关键区域丢失，影响训练数据的完整性。

常用场景

经典使用场景

在文本到图像生成领域，CyberHarem/ch_en_arknights 数据集为研究者提供了一个聚焦于《明日方舟》角色“陈”的高质量图像-标签配对资源。该数据集包含500张精心采集的图片，并附有详尽的标签描述，涵盖了角色的核心视觉特征如龙角、蓝发、长尾等。其最经典的应用场景是作为条件生成模型（例如Stable Diffusion）的微调训练集，通过标签引导模型精准学习特定角色的外观细节与风格化表达，从而在推理阶段实现从文本描述到角色专属图像的稳定生成。

解决学术问题

该数据集有效解决了动漫角色生成领域面临的细粒度概念对齐难题。在学术研究中，通用文本到图像模型常难以准确还原特定角色的标志性元素与服饰组合，而CyberHarem/ch_en_arknights通过结构化的多尺度图像裁剪（如480x480区域裁剪）与标签聚类分析，为模型提供了语义紧凑的视觉-文本对应关系。这推动了少样本角色定制生成、视觉概念解耦以及基于标签的细粒度图像编辑等研究方向的发展，为探索如何将虚拟角色知识高效迁移至生成模型提供了关键数据支撑。

实际应用

在实际应用中，该数据集被广泛用于动漫游戏产业的自动化内容创作与衍生品设计。通过微调基于该数据集的生成模型，用户能够快速生成符合角色设定的海报、同人插画或游戏内立绘变体，大幅降低美术资源的生产成本。此外，其提供的多分辨率版本（如800px与1200px）和三级裁剪数据，适配了从移动端预览到高清印刷的不同画质需求，使得角色素材的批量生成与风格迁移在商业流程中变得可落地、可复用。

数据集最近研究

最新研究方向

该数据集以《明日方舟》角色“陈”为核心，汇集了500张高质量图像及其精细标注的标签，涵盖了从原始高分辨率到多尺度裁剪的多种版本，为二次元角色生成和文本到图像（text-to-image）任务提供了标准化的训练资源。当前，该领域的前沿研究方向聚焦于利用此类精细标注的数据集，结合扩散模型（如Stable Diffusion）进行角色个性化生成、风格迁移以及多模态对齐。随着AIGC在游戏和动漫社区的快速普及，该数据集不仅支持了基于角色特征的零样本生成，还推动了标签聚类分析在服装、姿态等细粒度属性挖掘中的应用，为构建更可控、更符合用户偏好的生成模型奠定了数据基础，其影响延伸至数字内容创作和虚拟角色经济的蓬勃发展。