CyberHarem/kamiya_nao_idolmastercinderellagirls

这是一个名为kamiya_nao/神谷奈緒/카미야나오 (THE iDOLM@STER: Cinderella Girls)的数据集，包含500张图片及其标签。图片来源于多个网站，如danbooru、pixiv、zerochan等，并由DeepGHS团队的自动爬虫系统收集。数据集的核心标签包括brown_hair, long_hair, red_eyes, bangs, blunt_bangs, thick_eyebrows, breasts, hair_bun, single_hair_bun，这些标签在数据集中被修剪。README还提供了数据集的多个版本及其下载链接，以及如何使用waifuc加载原始数据集的代码示例。

This is the dataset of kamiya_nao/神谷奈緒/카미야나오 (THE iDOLM@STER: Cinderella Girls), containing 500 images and their tags. The core tags of this character include brown_hair, long_hair, red_eyes, bangs, blunt_bangs, thick_eyebrows, breasts, hair_bun, single_hair_bun, which are pruned in this dataset. The images are crawled from various sites such as danbooru, pixiv, and zerochan, and the auto-crawling system is powered by the DeepGHS Team. The dataset includes multiple packages with varying sizes and types, such as raw data, images with text, and cropped datasets. Additionally, the README provides a list of clusters showing tag clustering results for further analysis.

原始信息汇总

数据集概述

数据集名称

Dataset of kamiya_nao/神谷奈緒/카미야나오 (THE iDOLM@STER: Cinderella Girls)

数据集描述

该数据集包含500张图像及其标签，主题为kamiya_nao/神谷奈緒/카미야나오（《偶像大师：灰姑娘女孩》中的角色）。图像主要标签包括brown_hair, long_hair, red_eyes, bangs, blunt_bangs, thick_eyebrows, breasts, hair_bun, single_hair_bun。

数据来源

图像从多个网站（如danbooru, pixiv, zerochan等）爬取，爬虫系统由DeepGHS Team提供支持。

数据集包列表

名称	图像数量	大小	类型	描述
raw	500	691.78 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	500	396.16 MiB	IMG+TXT	短边不超过800像素的数据集。
stage3-p480-800	1225	846.19 MiB	IMG+TXT	3阶段裁剪数据集，区域不小于480x480像素。
1200	500	609.34 MiB	IMG+TXT	短边不超过1200像素的数据集。
stage3-p480-1200	1225	1.18 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	9		1girl, smile, solo, looking_at_viewer, belt, blush, earrings, navel, white_shorts, coat, midriff, open_mouth, bow, frills, hair_ornament, long_sleeves, short_shorts, white_background, black_thighhighs, holding_microphone, idol
1	39		1girl, solo, blush, looking_at_viewer, simple_background, white_background, white_shirt, school_uniform, blue_necktie, braid, long_sleeves, striped_necktie, plaid_skirt, pleated_skirt, upper_body, blue_jacket, smile
2	7		1girl, looking_at_viewer, open_mouth, smile, solo, blush, hair_flower, fingerless_gloves, thighhighs, skirt, microphone
3	10		1girl, elbow_gloves, midriff, skirt, solo, smile, belt, navel, hairband, microphone, open_mouth, black_gloves, blush, looking_at_viewer
4	11		1girl, blush, nipples, solo, looking_at_viewer, female_pubic_hair, medium_breasts, navel, large_breasts, completely_nude, sitting, sweat
5	5		1girl, blush, looking_at_viewer, open_mouth, solo, wet_shirt, bracelet, see-through, simple_background, white_background, white_shirt, bikini_skirt, low_twintails, navel, purple_bikini, short_sleeves, bikini_under_clothes, cowboy_shot, shirt_lift, smile
6	18		1girl, maid_headdress, blush, enmaided, solo, looking_at_viewer, frills, wrist_cuffs, maid_apron, thighhighs, bow, open_mouth, puffy_sleeves, short_sleeves
7	7		1girl, looking_at_viewer, navel, solo, blush, cleavage, collarbone, large_breasts, open_mouth, thighs, black_bikini, elbow_gloves, simple_background, white_background, bare_shoulders, black_gloves, black_thighhighs, micro_bikini, side-tie_bikini_bottom, black_choker

表格版本

#	样本数量	图像示例	1girl	smile	solo	looking_at_viewer	belt	blush	earrings	navel	white_shorts	coat	midriff	open_mouth	bow	frills	hair_ornament	long_sleeves	short_shorts	white_background	black_thighhighs	holding_microphone	idol	simple_background	white_shirt	school_uniform	blue_necktie	braid	striped_necktie	plaid_skirt	pleated_skirt	upper_body	blue_jacket	hair_flower	fingerless_gloves	thighhighs	skirt	microphone	elbow_gloves	hairband	black_gloves	nipples	female_pubic_hair	medium_breasts	large_breasts	completely_nude	sitting	sweat	wet_shirt	bracelet	see-through	bikini_skirt	low_twintails	purple_bikini	short_sleeves	bikini_under_clothes	cowboy_shot	shirt_lift	maid_headdress	enmaided	wrist_cuffs	maid_apron	puffy_sleeves	cleavage	collarbone	thighs	black_bikini	bare_shoulders	micro_bikini	side-tie_bikini_bottom	black_choker
0	9		X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X	X

搜集汇总

背景与挑战

背景概述

该数据集专注于文本到图像生成任务，规模在1千到1万条数据之间，标签包括艺术和敏感内容警告，使用MIT许可证。数据集主要用于生成与特定角色或内容相关的图像，适合艺术创作和敏感内容处理的研究。