CyberHarem/kousaka_reina_soundeuphonium

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

# Dataset of Kousaka Reina/高坂麗奈 (Sound! Euphonium)

This is the dataset of Kousaka Reina/高坂麗奈 (Sound! Euphonium), containing 403 images and their tags.

The core tags of this character are `black_hair, long_hair, purple_eyes`, 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 | 403 | 271.59 MiB | [Download](https://huggingface.co/datasets/CyberHarem/kousaka_reina_soundeuphonium/resolve/main/dataset-raw.zip) | Waifuc-Raw | Raw data with meta information (min edge aligned to 1400 if larger). |
| 1200 | 403 | 271.45 MiB | [Download](https://huggingface.co/datasets/CyberHarem/kousaka_reina_soundeuphonium/resolve/main/dataset-1200.zip) | IMG+TXT | dataset with the shorter side not exceeding 1200 pixels. |
| stage3-p480-1200 | 780 | 466.85 MiB | [Download](https://huggingface.co/datasets/CyberHarem/kousaka_reina_soundeuphonium/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/kousaka_reina_soundeuphonium',
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 | 6 |  |  |  |  |  | 1girl, blue_sailor_collar, blush, closed_mouth, kitauji_high_school_uniform, pink_neckerchief, serafuku, solo, white_shirt, upper_body, profile, short_sleeves, from_side |
| 1 | 6 |  |  |  |  |  | 1girl, kitauji_high_school_uniform, red_neckerchief, serafuku, solo, trumpet, playing_instrument, ponytail, holding_instrument, upper_body |
| 2 | 34 |  |  |  |  |  | 1girl, kitauji_high_school_uniform, serafuku, solo, red_neckerchief, brown_shirt, closed_mouth, white_sailor_collar, blush, upper_body, looking_at_viewer, long_sleeves, smile |
| 3 | 6 |  |  |  |  |  | 1girl, blush, closed_mouth, portrait, solo, blurry, kitauji_high_school_uniform, looking_at_viewer |
| 4 | 5 |  |  |  |  |  | 1girl, blush, close-up, closed_mouth, solo, looking_at_viewer, portrait |
| 5 | 7 |  |  |  |  |  | 1girl, brown_skirt, kitauji_high_school_uniform, pleated_skirt, red_neckerchief, solo, long_sleeves, sailor_collar, school_bag, brown_shirt, blush, brown_serafuku, standing |
| 6 | 8 |  |  |  |  |  | 2girls, blurry, blush, kitauji_high_school_uniform, serafuku, brown_hair, solo_focus, instrument, looking_at_another, sailor_collar, closed_mouth, neckerchief |
| 7 | 7 |  |  |  |  |  | 2girls, blush, looking_at_another, yuri, brown_hair, profile, close-up, closed_mouth, kitauji_high_school_uniform, blurry_background, from_side, short_hair |
| 8 | 14 |  |  |  |  |  | blue_kimono, yukata, blush, ponytail, 1girl, obi, solo, closed_mouth, hair_ornament |

### Table Version

| # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | 1girl | blue_sailor_collar | blush | closed_mouth | kitauji_high_school_uniform | pink_neckerchief | serafuku | solo | white_shirt | upper_body | profile | short_sleeves | from_side | red_neckerchief | trumpet | playing_instrument | ponytail | holding_instrument | brown_shirt | white_sailor_collar | looking_at_viewer | long_sleeves | smile | portrait | blurry | close-up | brown_skirt | pleated_skirt | sailor_collar | school_bag | brown_serafuku | standing | 2girls | brown_hair | solo_focus | instrument | looking_at_another | neckerchief | yuri | blurry_background | short_hair | blue_kimono | yukata | obi | hair_ornament |
|----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------|:---------------------|:--------|:---------------|:------------------------------|:-------------------|:-----------|:-------|:--------------|:-------------|:----------|:----------------|:------------|:------------------|:----------|:---------------------|:-----------|:---------------------|:--------------|:----------------------|:--------------------|:---------------|:--------|:-----------|:---------|:-----------|:--------------|:----------------|:----------------|:-------------|:-----------------|:-----------|:---------|:-------------|:-------------|:-------------|:---------------------|:--------------|:-------|:--------------------|:-------------|:--------------|:---------|:------|:----------------|
| 0 | 6 |  |  |  |  |  | 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 | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| 2 | 34 |  |  |  |  |  | X | | X | X | X | | X | X | | X | | | | X | | | | | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | |
| 3 | 6 |  |  |  |  |  | X | | X | X | X | | | X | | | | | | | | | | | | | X | | | X | X | | | | | | | | | | | | | | | | | | | | |
| 4 | 5 |  |  |  |  |  | X | | X | X | | | | X | | | | | | | | | | | | | X | | | X | | X | | | | | | | | | | | | | | | | | | | |
| 5 | 7 |  |  |  |  |  | X | | X | | X | | | X | | | | | | X | | | | | X | | | X | | | | | X | X | X | X | X | X | | | | | | | | | | | | | |
| 6 | 8 |  |  |  |  |  | | | X | X | X | | X | | | | | | | | | | | | | | | | | | X | | | | X | | | | X | X | X | X | X | X | | | | | | | |
| 7 | 7 |  |  |  |  |  | | | X | X | X | | | | | | X | | X | | | | | | | | | | | | | X | | | | | | | X | X | | | X | | X | X | X | | | | |
| 8 | 14 |  |  |  |  |  | X | | X | X | | | | X | | | | | | | | | X | | | | | | | | | | | | | | | | | | | | | | | | | X | X | X | X |

原始信息汇总

高坂麗奈（吹響吧！上低音號）數據集

數據集概述

• 名稱: 高坂麗奈（吹響吧！上低音號）數據集
• 圖像數量: 403張
• 核心標籤: black_hair, long_hair, purple_eyes
• 標籤類型: 文本標籤
• 圖像來源: 從多個網站爬取，如danbooru, pixiv, zerochan等

數據集包列表

名稱	圖像數量	大小	類型	描述
raw	403	271.59 MiB	Waifuc-Raw	原始數據，包含元信息（最小邊對齊到1400像素）
1200	403	271.45 MiB	IMG+TXT	短邊不超過1200像素的數據集
stage3-p480-1200	780	466.85 MiB	IMG+TXT	3階段裁剪數據集，區域不小於480x480像素

標籤聚類結果

原始文本版本

#	樣本數	圖像1	圖像2	圖像3	圖像4	圖像5	標籤
0	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, blue_sailor_collar, blush, closed_mouth, kitauji_high_school_uniform, pink_neckerchief, serafuku, solo, white_shirt, upper_body, profile, short_sleeves, from_side
1	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, kitauji_high_school_uniform, red_neckerchief, serafuku, solo, trumpet, playing_instrument, ponytail, holding_instrument, upper_body
2	34	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, kitauji_high_school_uniform, serafuku, solo, red_neckerchief, brown_shirt, closed_mouth, white_sailor_collar, blush, upper_body, looking_at_viewer, long_sleeves, smile
3	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, blush, closed_mouth, portrait, solo, blurry, kitauji_high_school_uniform, looking_at_viewer
4	5	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, blush, close-up, closed_mouth, solo, looking_at_viewer, portrait
5	7	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	1girl, brown_skirt, kitauji_high_school_uniform, pleated_skirt, red_neckerchief, solo, long_sleeves, sailor_collar, school_bag, brown_shirt, blush, brown_serafuku, standing
6	8	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	2girls, blurry, blush, kitauji_high_school_uniform, serafuku, brown_hair, solo_focus, instrument, looking_at_another, sailor_collar, closed_mouth, neckerchief
7	7	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	2girls, blush, looking_at_another, yuri, brown_hair, profile, close-up, closed_mouth, kitauji_high_school_uniform, blurry_background, from_side, short_hair
8	14	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	blue_kimono, yukata, blush, ponytail, 1girl, obi, solo, closed_mouth, hair_ornament

表格版本

#	樣本數	圖像1	圖像2	圖像3	圖像4	圖像5	1girl	blue_sailor_collar	blush	closed_mouth	kitauji_high_school_uniform	pink_neckerchief	serafuku	solo	white_shirt	upper_body	profile	short_sleeves	from_side	red_neckerchief	trumpet	playing_instrument	ponytail	holding_instrument	brown_shirt	white_sailor_collar	looking_at_viewer	long_sleeves	smile	portrait	blurry	close-up	brown_skirt	pleated_skirt	sailor_collar	school_bag	brown_serafuku	standing	2girls	brown_hair	solo_focus	instrument	looking_at_another	neckerchief	yuri	blurry_background	short_hair	blue_kimono	yukata	obi	hair_ornament
0	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X	X	X	X	X	X	X	X	X	X	X	X	X								X	X
1	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X				X		X	X		X				X	X	X	X	X				X
2	34	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X		X	X	X		X	X		X				X					X	X	X	X	X
3	6	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X		X	X	X			X													X			X	X
4	5	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X		X	X				X													X			X		X
5	7	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X		X		X			X						X					X			X					X	X	X	X	X	X
6	8	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]			X	X	X		X																		X				X				X	X	X	X	X	X
7	7	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]			X	X	X						X		X													X							X	X			X		X	X	X
8	14	[圖像]	[圖像]	[圖像]	[圖像]	[圖像]	X		X	X				X									X																									X	X	X	X

搜集汇总

数据集介绍

构建方式

在动漫角色图像数据集构建领域，针对《吹响吧！上低音号》中的高坂丽奈这一角色，CyberHarem团队构建了该数据集。数据集的构建依托于DeepGHS团队开发的自动化爬取系统，从Danbooru、Pixiv、Zerochan等多个图像托管平台收集原始图像。收集到的403张图像经过筛选，并剔除了角色核心标签（黑发、长发、紫瞳）以简化标注。数据集提供了多种处理版本：原始版（raw）保留元信息并将短边对齐至1400像素；1200版将图像短边限制在1200像素以内；stage3-p480-1200版则通过三阶段裁剪，确保每张图像的有效区域不小于480×480像素，从而适配不同训练需求。

使用方法

数据集的使用方式灵活，支持通过waifuc库进行加载。用户可首先利用huggingface_hub下载原始压缩包，解压至本地目录后，借助LocalSource接口迭代访问图像及其元数据（如文件名和标签）。对于需要标准尺寸输入的训练任务，可直接使用1200或stage3-p480-1200版本的ZIP文件。此外，数据集的标签聚类结果以表格形式提供，可作为先验知识辅助模型设计或数据增强策略的制定，尤其适用于文本到图像生成模型的微调与评估。

背景与挑战

背景概述

在动漫文化蓬勃发展的背景下，角色图像数据集成为文本到图像生成模型训练的关键资源。CyberHarem团队于近期发布了高坂麗奈（出自《吹响吧！上低音号》）的专用图像数据集，该数据集由DeepGHS团队开发的自动化爬取系统从Danbooru、Pixiv、Zerochan等多个平台收集，共包含403张经过标注的图像。核心研究问题聚焦于如何为特定动漫角色构建高质量、多视角且标签完备的训练数据，以支撑基于扩散模型的角色生成任务。该数据集通过提供原始图像、统一尺寸版本及多阶段裁剪版本，显著降低了研究者获取特定角色训练数据的门槛，在二次元图像生成领域展现出重要的应用潜力。

当前挑战

当前数据集面临的核心挑战在于角色多样性与数据规模的平衡。尽管高坂麗奈具有黑发、长发、紫眼等显著特征，但403张图像的数量限制了模型对角色在不同场景、服饰及表情下的泛化能力。数据构建过程中，自动化爬取虽提升了效率，却引入了标签噪声与版权风险，部分图像来自不同平台导致风格差异显著，增加了模型学习的难度。此外，数据集中仅包含单一角色的图像，缺乏与其他角色交互的复杂场景，使得模型在生成多角色互动或特定剧情情境下的图像时表现欠佳。这些挑战共同制约了数据集在精细化角色生成任务中的实际效能。

常用场景

经典使用场景

在动漫角色生成与定制化图像合成领域，CyberHarem/kousaka_reina_soundeuphonium 数据集为基于文本到图像的生成模型提供了高质量、标签化的训练素材。该数据集聚焦于《吹响吧！上低音号》中的高坂丽奈这一角色，包含403张经过精细标注的图像，核心标签如黑发、长发、紫瞳等被系统化修剪，便于模型精准捕捉角色特征。研究者常利用此数据集进行角色一致性生成任务的训练，通过Waifuc等框架加载原始数据，实现从文本描述到稳定角色形象的映射，是动漫风格迁移与个性化生成研究的重要基石。

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泛化之间的学术难题。传统生成模型在处理特定角色时，常因训练数据不足或标签噪声导致特征混淆，而高坂丽奈数据集通过多源爬取（如Danbooru、Pixiv）和标签聚类（如校服、乐器场景的细粒度分组），为研究提供了结构化的特征空间。它支持学界探索条件生成中的标签解耦、小样本学习下的角色重建，以及跨模态对齐等核心问题，推动了可控图像生成理论在二次元领域的实证进展。

实际应用

在实际应用中，该数据集赋能了虚拟偶像内容生产与动漫衍生品设计等场景。通过训练定制化的文本到图像模型，创作者可快速生成高坂丽奈在特定服饰（如北宇治高中制服、浴衣）或动作（如吹奏小号）下的高质量图像，大幅降低手工绘制成本。在游戏开发、动画前期概念设计以及粉丝二创社区中，该数据集为角色IP的数字化资产管理提供了标准化的数据支撑，促进了从静态图库到动态生成管线的工业级转化。

数据集最近研究

最新研究方向

在文本到图像生成领域，高质量、细粒度的角色数据集正成为推动动漫风格图像生成模型发展的关键资源。以《吹响吧！上低音号》中高坂丽奈这一角色为对象的数据集，通过整合来自Danbooru、Pixiv等平台的403张标注图像，并采用自动爬取与多阶段裁剪技术，构建了包含原始数据、标准化尺寸及三级裁剪版本的结构化资源。这一工作呼应了当前二次元角色定制化生成的热潮，尤其是在LoRA微调与Stable Diffusion等扩散模型的应用中，精细化的角色标签（如发色、服饰、场景聚类）能够显著提升生成图像的风格一致性与角色保真度。该数据集不仅为动漫角色生成提供了标准化的训练素材，更通过标签聚类分析揭示了角色在不同着装与场景下的视觉模式，为跨模态生成中的语义对齐研究提供了基准。其开源特性与waifuc框架的兼容性，进一步降低了研究门槛，推动了社区驱动的角色生成技术迭代，对个性化虚拟角色创作与动漫IP数字化具有重要的实践意义。