CyberHarem/eponine_fireemblem

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

# Dataset of eponine (Fire Emblem)

This is the dataset of eponine (Fire Emblem), containing 366 images and their tags.

The core tags of this character are `braid, ahoge, twin_braids, bangs, long_hair, blue_eyes, breasts, twintails, parted_bangs, low_twintails, white_hair, hairband, medium_breasts, white_hairband`, 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 | 366 | 422.92 MiB | [Download](https://huggingface.co/datasets/CyberHarem/eponine_fireemblem/resolve/main/dataset-raw.zip) | Waifuc-Raw | Raw data with meta information (min edge aligned to 1400 if larger). |
| 800 | 366 | 238.45 MiB | [Download](https://huggingface.co/datasets/CyberHarem/eponine_fireemblem/resolve/main/dataset-800.zip) | IMG+TXT | dataset with the shorter side not exceeding 800 pixels. |
| stage3-p480-800 | 769 | 475.22 MiB | [Download](https://huggingface.co/datasets/CyberHarem/eponine_fireemblem/resolve/main/dataset-stage3-p480-800.zip) | IMG+TXT | 3-stage cropped dataset with the area not less than 480x480 pixels. |
| 1200 | 366 | 368.23 MiB | [Download](https://huggingface.co/datasets/CyberHarem/eponine_fireemblem/resolve/main/dataset-1200.zip) | IMG+TXT | dataset with the shorter side not exceeding 1200 pixels. |
| stage3-p480-1200 | 769 | 689.01 MiB | [Download](https://huggingface.co/datasets/CyberHarem/eponine_fireemblem/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/eponine_fireemblem',
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 | 18 |  |  |  |  |  | 1girl, solo, looking_at_viewer, low_twin_braids, hair_flower, wings, official_alternate_costume, open_mouth, upper_body, cleavage, smile |
| 1 | 35 |  |  |  |  |  | 1girl, hooded_capelet, upper_body, choker, chest_harness, solo, buttons, turtleneck, hood_down, closed_mouth, looking_at_viewer, red_hood, cleavage, leather, smile, red_capelet |
| 2 | 22 |  |  |  |  |  | 1girl, hooded_capelet, belt, chest_harness, solo, buttons, choker, holding_bow_(weapon), hood_down, leather_boots, looking_at_viewer, arrow_(projectile), red_hood, key, low_twin_braids, red_capelet, full_body |
| 3 | 7 |  |  |  |  |  | 1girl, cleavage, earrings, hood_up, hooded_cape, looking_at_viewer, official_alternate_costume, turtleneck, mask, open_mouth, solo, upper_body, black_cape, black_gloves, blue_bodysuit, feathers, o-ring, smile, upper_teeth_only, holding, one_eye_closed, chest_harness, covered_navel, rose, skin_tight |
| 4 | 7 |  |  |  |  |  | 1girl, black_footwear, blue_bodysuit, earrings, holding, looking_at_viewer, o-ring, official_alternate_costume, thigh_boots, turtleneck, black_cape, black_gloves, chest_harness, cleavage, hooded_cape, leather_belt, mask, skin_tight, solo, latex_bodysuit, blue_ribbon, covered_navel, hair_ribbon, key, hood_up, red_cape, satchel, tongue_out |
| 5 | 5 |  |  |  |  |  | 1girl, alternate_costume, navel, solo, bikini, looking_at_viewer, cleavage, hair_flower, jewelry, holding, midriff |

### Table Version

| # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | 1girl | solo | looking_at_viewer | low_twin_braids | hair_flower | wings | official_alternate_costume | open_mouth | upper_body | cleavage | smile | hooded_capelet | choker | chest_harness | buttons | turtleneck | hood_down | closed_mouth | red_hood | leather | red_capelet | belt | holding_bow_(weapon) | leather_boots | arrow_(projectile) | key | full_body | earrings | hood_up | hooded_cape | mask | black_cape | black_gloves | blue_bodysuit | feathers | o-ring | upper_teeth_only | holding | one_eye_closed | covered_navel | rose | skin_tight | black_footwear | thigh_boots | leather_belt | latex_bodysuit | blue_ribbon | hair_ribbon | red_cape | satchel | tongue_out | alternate_costume | navel | bikini | jewelry | midriff |
|----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------|:-------|:--------------------|:------------------|:--------------|:--------|:-----------------------------|:-------------|:-------------|:-----------|:--------|:-----------------|:---------|:----------------|:----------|:-------------|:------------|:---------------|:-----------|:----------|:--------------|:-------|:-----------------------|:----------------|:---------------------|:------|:------------|:-----------|:----------|:--------------|:-------|:-------------|:---------------|:----------------|:-----------|:---------|:-------------------|:----------|:-----------------|:----------------|:-------|:-------------|:-----------------|:--------------|:---------------|:-----------------|:--------------|:--------------|:-----------|:----------|:-------------|:--------------------|:--------|:---------|:----------|:----------|
| 0 | 18 |  |  |  |  |  | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 1 | 35 |  |  |  |  |  | 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 | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 3 | 7 |  |  |  |  |  | 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 | 7 |  |  |  |  |  | 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 |

原始信息汇总

数据集概述

数据集信息

• 名称: Dataset of eponine (Fire Emblem)
• 许可证: MIT
• 任务类别: text-to-image
• 标签: art, not-for-all-audiences
• 大小类别: n<1K

数据集内容

• 图像数量: 366张
• 核心标签: braid, ahoge, twin_braids, bangs, long_hair, blue_eyes, breasts, twintails, parted_bangs, low_twintails, white_hair, hairband, medium_breasts, white_hairband

数据包列表

名称	图像数量	大小	类型	描述
raw	366	422.92 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	366	238.45 MiB	IMG+TXT	短边不超过800像素的数据集。
stage3-p480-800	769	475.22 MiB	IMG+TXT	3阶段裁剪数据集，区域不小于480x480像素。
1200	366	368.23 MiB	IMG+TXT	短边不超过1200像素的数据集。
stage3-p480-1200	769	689.01 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

#	Samples	Img-1	Img-2	Img-3	Img-4	Img-5	标签
0	18						1girl, solo, looking_at_viewer, low_twin_braids, hair_flower, wings, official_alternate_costume, open_mouth, upper_body, cleavage, smile
1	35						1girl, hooded_capelet, upper_body, choker, chest_harness, solo, buttons, turtleneck, hood_down, closed_mouth, looking_at_viewer, red_hood, cleavage, leather, smile, red_capelet
2	22						1girl, hooded_capelet, belt, chest_harness, solo, buttons, choker, holding_bow_(weapon), hood_down, leather_boots, looking_at_viewer, arrow_(projectile), red_hood, key, low_twin_braids, red_capelet, full_body
3	7						1girl, cleavage, earrings, hood_up, hooded_cape, looking_at_viewer, official_alternate_costume, turtleneck, mask, open_mouth, solo, upper_body, black_cape, black_gloves, blue_bodysuit, feathers, o-ring, smile, upper_teeth_only, holding, one_eye_closed, chest_harness, covered_navel, rose, skin_tight
4	7						1girl, black_footwear, blue_bodysuit, earrings, holding, looking_at_viewer, o-ring, official_alternate_costume, thigh_boots, turtleneck, black_cape, black_gloves, chest_harness, cleavage, hooded_cape, leather_belt, mask, skin_tight, solo, latex_bodysuit, blue_ribbon, covered_navel, hair_ribbon, key, hood_up, red_cape, satchel, tongue_out
5	5						1girl, alternate_costume, navel, solo, bikini, looking_at_viewer, cleavage, hair_flower, jewelry, holding, midriff

表格版本

#	Samples	Img-1	Img-2	Img-3	Img-4	Img-5	1girl	solo	looking_at_viewer	low_twin_braids	hair_flower	wings	official_alternate_costume	open_mouth	upper_body	cleavage	smile	hooded_capelet	choker	chest_harness	buttons	turtleneck	hood_down	closed_mouth	red_hood	leather	red_capelet	belt	holding_bow_(weapon)	leather_boots	arrow_(projectile)	key	full_body	earrings	hood_up	hooded_cape	mask	black_cape	black_gloves	blue_bodysuit	feathers	o-ring	upper_teeth_only	holding	one_eye_closed	covered_navel	rose	skin_tight	black_footwear	thigh_boots	leather_belt	latex_bodysuit	blue_ribbon	hair_ribbon	red_cape	satchel	tongue_out	alternate_costume	navel	bikini	jewelry	midriff
0	18						X	X	X	X	X	X	X	X	X	X	X
1	35						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	X
3	7						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	7						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

搜集汇总

数据集介绍

构建方式

在动漫与游戏角色图像生成领域，高质量、结构化的数据集对于模型微调至关重要。该数据集聚焦于《火焰之纹章》中的角色eponine，通过自动化爬虫系统从Danbooru、Pixiv、Zerochan等多源平台采集了366张图像及其标签。数据集的构建遵循了精细化的处理流程：首先，对原始图像进行元信息保留并统一将短边对齐至1400像素；随后，生成不同分辨率版本（800与1200像素），并进一步采用三阶段裁剪策略，产出面积不低于480×480像素的增强数据集。最终，数据集以多种压缩包形式发布，便于用户按需选用。

使用方法

用户可通过Hugging Face Hub便捷地获取该数据集。推荐使用waifuc库加载原始数据集：首先利用huggingface_hub下载dataset-raw.zip压缩包，解压至本地目录后，通过LocalSource接口即可逐项读取图像及其标签元数据。对于不同应用场景，用户可选择800、1200或裁剪版数据集，这些版本均以IMG+TXT格式存储，可直接用于文本到图像生成模型的训练。代码示例已在文档中提供，确保了使用的低门槛与可复现性。

背景与挑战

背景概述

在二次元文化蓬勃发展的浪潮中，角色图像数据集成为文本到图像生成模型训练的关键基石。CyberHarem/eponine_fireemblem数据集由DeepGHS团队于近年创建，专注于收录《火焰之纹章》系列中角色艾波妮的视觉素材。该数据集核心研究问题在于如何通过多源爬取与自动化标注技术，构建高质量、细粒度的角色图像库，以支撑动漫风格的人物生成任务。数据集包含366张经过严格筛选的图像及其标签，覆盖角色标志性的双辫、刘海、白色长发等视觉特征，为领域内研究者提供了标准化的训练资源，对推动二次元角色定制化生成技术的发展具有显著影响力。

当前挑战

该数据集面临多重挑战。在领域问题层面，文本到图像生成任务要求模型精准理解并还原角色复杂的外观特征（如发型、服饰、配饰的组合），而二次元角色风格多变，同一角色在不同作品或画师笔下存在显著差异，增加了模型泛化的难度。在构建过程中，数据爬取自Danbooru、Pixiv等多个站点，面临来源异构、标签不一致及版权合规性问题；同时，图像质量参差不齐，需通过边缘对齐、裁剪等预处理流程进行标准化，而小规模数据集（n<1K）又容易导致模型过拟合，难以捕捉角色的全部视觉语义。

常用场景

经典使用场景

在文本到图像生成领域，CyberHarem/eponine_fireemblem数据集为角色一致性生成任务提供了宝贵的资源。该数据集汇集了来自Danbooru、Pixiv等平台的366张高质量图像，并附有精细的标签注释，核心标签涵盖发型、瞳色、服饰等关键特征。研究者常基于此数据集微调扩散模型，如Stable Diffusion，以实现对《火焰纹章》中艾波宁角色的精准复现，确保生成图像在风格、姿态和细节上保持高度一致。

解决学术问题

该数据集有效解决了动漫角色生成中身份特征保持与风格迁移的学术难题。传统生成模型常因缺乏特定角色的细粒度标注数据，导致生成结果出现特征混淆或风格飘移。通过提供多角度、多服饰变体的标注图像，并引入三级裁剪预处理以优化训练效率，该数据集为少样本角色定制化生成提供了标准化基准，推动了可控生成、概念学习等研究方向的发展。

实际应用

在实际应用中，该数据集赋能游戏与动漫产业的自动化内容生产。开发者可基于此数据集训练定制化模型，快速生成角色立绘、宣传海报或同人创作素材，大幅降低人工绘制成本。同时，其标注的服饰聚类结果（如官方换装、泳装等变体）为虚拟偶像运营和衍生品设计提供了数据支撑，助力实现从概念草图到成品图像的工业化流水线。

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，随着文本到图像（text-to-image）技术的迅猛发展，高质量、细粒度的角色数据集成为推动模型精准刻画特定人物形象的关键基石。CyberHarem/eponine_fireemblem 数据集聚焦于《火焰之纹章》中的角色艾波妮，收录了366张经过精细标签标注的图像，其核心标签涵盖发型、瞳色、服饰等多维度特征，为角色定制化生成提供了丰富的语义锚点。该数据集的前沿研究方向集中于利用多阶段裁剪与变体增强策略提升生成模型的泛化能力，例如通过stage3-p480-800等不同分辨率版本探索细节保留与计算效率的平衡。此外，标签聚类分析揭示了角色在不同服装与场景下的视觉模式，为可控图像生成中的风格迁移与属性解耦研究注入了新的可能性。这一工作不仅服务于游戏同人创作生态，更对推动个性化AI艺术创作与虚拟角色工业化生产具有深远意义。