five

CyberHarem/mikazuki_kantaicollection

收藏
Hugging Face2024-01-16 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/CyberHarem/mikazuki_kantaicollection
下载链接
链接失效反馈
官方服务:
资源简介:
--- license: mit task_categories: - text-to-image tags: - art - not-for-all-audiences size_categories: - n<1K --- # Dataset of mikazuki (Kantai Collection) This is the dataset of mikazuki (Kantai Collection), containing 336 images and their tags. The core tags of this character are `long_hair, black_hair, ahoge, yellow_eyes, hair_between_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 | 336 | 236.44 MiB | [Download](https://huggingface.co/datasets/CyberHarem/mikazuki_kantaicollection/resolve/main/dataset-raw.zip) | Waifuc-Raw | Raw data with meta information (min edge aligned to 1400 if larger). | | 800 | 336 | 167.29 MiB | [Download](https://huggingface.co/datasets/CyberHarem/mikazuki_kantaicollection/resolve/main/dataset-800.zip) | IMG+TXT | dataset with the shorter side not exceeding 800 pixels. | | stage3-p480-800 | 736 | 345.20 MiB | [Download](https://huggingface.co/datasets/CyberHarem/mikazuki_kantaicollection/resolve/main/dataset-stage3-p480-800.zip) | IMG+TXT | 3-stage cropped dataset with the area not less than 480x480 pixels. | | 1200 | 336 | 223.33 MiB | [Download](https://huggingface.co/datasets/CyberHarem/mikazuki_kantaicollection/resolve/main/dataset-1200.zip) | IMG+TXT | dataset with the shorter side not exceeding 1200 pixels. | | stage3-p480-1200 | 736 | 434.18 MiB | [Download](https://huggingface.co/datasets/CyberHarem/mikazuki_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/mikazuki_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 | 12 | ![](samples/0/clu0-sample0.png) | ![](samples/0/clu0-sample1.png) | ![](samples/0/clu0-sample2.png) | ![](samples/0/clu0-sample3.png) | ![](samples/0/clu0-sample4.png) | 1girl, black_serafuku, long_sleeves, looking_at_viewer, solo, skirt, blush, crescent, necktie, white_background, animal_ears, kemonomimi_mode, open_mouth, tail | | 1 | 40 | ![](samples/1/clu1-sample0.png) | ![](samples/1/clu1-sample1.png) | ![](samples/1/clu1-sample2.png) | ![](samples/1/clu1-sample3.png) | ![](samples/1/clu1-sample4.png) | 1girl, black_serafuku, solo, crescent_pin, looking_at_viewer, long_sleeves, white_necktie, simple_background, white_background, black_skirt, black_sailor_collar, blush, open_mouth, smile, pleated_skirt, white_neckerchief, twitter_username, black_shirt, one-hour_drawing_challenge | | 2 | 12 | ![](samples/2/clu2-sample0.png) | ![](samples/2/clu2-sample1.png) | ![](samples/2/clu2-sample2.png) | ![](samples/2/clu2-sample3.png) | ![](samples/2/clu2-sample4.png) | detached_collar, looking_at_viewer, playboy_bunny, rabbit_ears, wrist_cuffs, 1girl, solo, strapless_leotard, black_leotard, fake_animal_ears, alternate_costume, crescent, red_bowtie, small_breasts, open_mouth, black_pantyhose, cowboy_shot, flat_chest, rabbit_tail, white_background | | 3 | 13 | ![](samples/3/clu3-sample0.png) | ![](samples/3/clu3-sample1.png) | ![](samples/3/clu3-sample2.png) | ![](samples/3/clu3-sample3.png) | ![](samples/3/clu3-sample4.png) | 1girl, looking_at_viewer, solo, cowboy_shot, simple_background, collarbone, flat_chest, smile, blue_one-piece_swimsuit, blush, old_school_swimsuit, white_background, open_mouth, ass_visible_through_thighs, gradient_background, twitter_username, covered_navel, grey_background | | 4 | 5 | ![](samples/4/clu4-sample0.png) | ![](samples/4/clu4-sample1.png) | ![](samples/4/clu4-sample2.png) | ![](samples/4/clu4-sample3.png) | ![](samples/4/clu4-sample4.png) | 1girl, barefoot, blue_one-piece_swimsuit, blush, collarbone, competition_school_swimsuit, indoors, looking_at_viewer, poolside, rei_no_pool, sitting, solo, window, small_breasts, spread_legs, water, alternate_costume, covered_navel, smile, tile_floor, cameltoe, competition_swimsuit, feet, full_body, pool_ladder | | 5 | 7 | ![](samples/5/clu5-sample0.png) | ![](samples/5/clu5-sample1.png) | ![](samples/5/clu5-sample2.png) | ![](samples/5/clu5-sample3.png) | ![](samples/5/clu5-sample4.png) | 1girl, looking_at_viewer, simple_background, solo, white_background, black_dress, blush, enmaided, white_apron, frilled_apron, maid_apron, maid_headdress, puffy_sleeves, smile, black_footwear, bow, brown_eyes, cowboy_shot, dated, full_body, heart, shoes, short_sleeves | ### Table Version | # | Samples | Img-1 | Img-2 | Img-3 | Img-4 | Img-5 | 1girl | black_serafuku | long_sleeves | looking_at_viewer | solo | skirt | blush | crescent | necktie | white_background | animal_ears | kemonomimi_mode | open_mouth | tail | crescent_pin | white_necktie | simple_background | black_skirt | black_sailor_collar | smile | pleated_skirt | white_neckerchief | twitter_username | black_shirt | one-hour_drawing_challenge | detached_collar | playboy_bunny | rabbit_ears | wrist_cuffs | strapless_leotard | black_leotard | fake_animal_ears | alternate_costume | red_bowtie | small_breasts | black_pantyhose | cowboy_shot | flat_chest | rabbit_tail | collarbone | blue_one-piece_swimsuit | old_school_swimsuit | ass_visible_through_thighs | gradient_background | covered_navel | grey_background | barefoot | competition_school_swimsuit | indoors | poolside | rei_no_pool | sitting | window | spread_legs | water | tile_floor | cameltoe | competition_swimsuit | feet | full_body | pool_ladder | black_dress | enmaided | white_apron | frilled_apron | maid_apron | maid_headdress | puffy_sleeves | black_footwear | bow | brown_eyes | dated | heart | shoes | short_sleeves | |----:|----------:|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------------------------------|:--------|:-----------------|:---------------|:--------------------|:-------|:--------|:--------|:-----------|:----------|:-------------------|:--------------|:------------------|:-------------|:-------|:---------------|:----------------|:--------------------|:--------------|:----------------------|:--------|:----------------|:--------------------|:-------------------|:--------------|:-----------------------------|:------------------|:----------------|:--------------|:--------------|:--------------------|:----------------|:-------------------|:--------------------|:-------------|:----------------|:------------------|:--------------|:-------------|:--------------|:-------------|:--------------------------|:----------------------|:-----------------------------|:----------------------|:----------------|:------------------|:-----------|:------------------------------|:----------|:-----------|:--------------|:----------|:---------|:--------------|:--------|:-------------|:-----------|:-----------------------|:-------|:------------|:--------------|:--------------|:-----------|:--------------|:----------------|:-------------|:-----------------|:----------------|:-----------------|:------|:-------------|:--------|:--------|:--------|:----------------| | 0 | 12 | ![](samples/0/clu0-sample0.png) | ![](samples/0/clu0-sample1.png) | ![](samples/0/clu0-sample2.png) | ![](samples/0/clu0-sample3.png) | ![](samples/0/clu0-sample4.png) | X | X | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 1 | 40 | ![](samples/1/clu1-sample0.png) | ![](samples/1/clu1-sample1.png) | ![](samples/1/clu1-sample2.png) | ![](samples/1/clu1-sample3.png) | ![](samples/1/clu1-sample4.png) | X | X | X | X | X | | X | | | X | | | X | | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 2 | 12 | ![](samples/2/clu2-sample0.png) | ![](samples/2/clu2-sample1.png) | ![](samples/2/clu2-sample2.png) | ![](samples/2/clu2-sample3.png) | ![](samples/2/clu2-sample4.png) | X | | | X | X | | | X | | X | | | X | | | | | | | | | | | | | X | X | X | X | X | X | X | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 3 | 13 | ![](samples/3/clu3-sample0.png) | ![](samples/3/clu3-sample1.png) | ![](samples/3/clu3-sample2.png) | ![](samples/3/clu3-sample3.png) | ![](samples/3/clu3-sample4.png) | X | | | X | X | | X | | | X | | | X | | | | X | | | X | | | X | | | | | | | | | | | | | | X | X | | X | X | X | X | X | X | X | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4 | 5 | ![](samples/4/clu4-sample0.png) | ![](samples/4/clu4-sample1.png) | ![](samples/4/clu4-sample2.png) | ![](samples/4/clu4-sample3.png) | ![](samples/4/clu4-sample4.png) | 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 | 7 | ![](samples/5/clu5-sample0.png) | ![](samples/5/clu5-sample1.png) | ![](samples/5/clu5-sample2.png) | ![](samples/5/clu5-sample3.png) | ![](samples/5/clu5-sample4.png) | X | | | X | X | | X | | | X | | | | | | | X | | | X | | | | | | | | | | | | | | | | | X | | | | | | | | | | | | | | | | | | | | | | | X | | X | X | X | X | X | X | X | X | X | X | X | X | X | X |
提供机构:
CyberHarem
原始信息汇总

数据集概述

数据集基本信息

  • 名称: Dataset of mikazuki (Kantai Collection)
  • 许可: MIT
  • 任务类别: text-to-image
  • 标签: art, not-for-all-audiences
  • 大小类别: n<1K
  • 图像数量: 336
  • 核心标签: long_hair, black_hair, ahoge, yellow_eyes, hair_between_eyes

数据集包列表

名称 图像数量 大小 类型 描述
raw 336 236.44 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)
800 336 167.29 MiB IMG+TXT 短边不超过800像素的数据集
stage3-p480-800 736 345.20 MiB IMG+TXT 3阶段裁剪数据集,区域不小于480x480像素
1200 336 223.33 MiB IMG+TXT 短边不超过1200像素的数据集
stage3-p480-1200 736 434.18 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 图像示例 标签
0 12 1girl, black_serafuku, long_sleeves, looking_at_viewer, solo, skirt, blush, crescent, necktie, white_background, animal_ears, kemonomimi_mode, open_mouth, tail
1 40 1girl, black_serafuku, solo, crescent_pin, looking_at_viewer, long_sleeves, white_necktie, simple_background, white_background, black_skirt, black_sailor_collar, blush, open_mouth, smile, pleated_skirt, white_neckerchief, twitter_username, black_shirt, one-hour_drawing_challenge
2 12 detached_collar, looking_at_viewer, playboy_bunny, rabbit_ears, wrist_cuffs, 1girl, solo, strapless_leotard, black_leotard, fake_animal_ears, alternate_costume, crescent, red_bowtie, small_breasts, open_mouth, black_pantyhose, cowboy_shot, flat_chest, rabbit_tail, white_background
3 13 1girl, looking_at_viewer, solo, cowboy_shot, simple_background, collarbone, flat_chest, smile, blue_one-piece_swimsuit, blush, old_school_swimsuit, white_background, open_mouth, ass_visible_through_thighs, gradient_background, twitter_username, covered_navel, grey_background
4 5 1girl, barefoot, blue_one-piece_swimsuit, blush, collarbone, competition_school_swimsuit, indoors, looking_at_viewer, poolside, rei_no_pool, sitting, solo, window, small_breasts, spread_legs, water, alternate_costume, covered_navel, smile, tile_floor, cameltoe, competition_swimsuit, feet, full_body, pool_ladder
5 7 1girl, looking_at_viewer, simple_background, solo, white_background, black_dress, blush, enmaided, white_apron, frilled_apron, maid_apron, maid_headdress, puffy_sleeves, smile, black_footwear, bow, brown_eyes, cowboy_shot, dated, full_body, heart, shoes, short_sleeves

表格版本

# 样本数量 图像示例 1girl black_serafuku long_sleeves looking_at_viewer solo skirt blush crescent necktie white_background animal_ears kemonomimi_mode open_mouth tail crescent_pin white_necktie simple_background black_skirt black_sailor_collar smile pleated_skirt white_neckerchief twitter_username black_shirt one-hour_drawing_challenge detached_collar playboy_bunny rabbit_ears wrist_cuffs strapless_leotard black_leotard fake_animal_ears alternate_costume red_bowtie small_breasts black_pantyhose cowboy_shot flat_chest rabbit_tail collarbone blue_one-piece_swimsuit old_school_swimsuit ass_visible_through_thighs gradient_background covered_navel grey_background barefoot competition_school_swimsuit indoors poolside rei_no_pool sitting window spread_legs water tile_floor cameltoe competition_swimsuit feet full_body pool_ladder black_dress enmaided white_apron frilled_apron maid_apron maid_headdress puffy_sleeves black_footwear bow brown_eyes dated heart shoes short_sleeves
0 12 X X X X X X X X X X X X X X
搜集汇总
数据集介绍
main_image_url
构建方式
在数字艺术与动漫文化交融的背景下,该数据集聚焦于《舰队Collection》中的角色“三日月”,系统性地收集了336幅图像及其对应的标签信息。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等多个知名图像平台广泛采集素材。为提升数据质量,核心标签如长髮、黑髮、呆毛、黄瞳、眼中髮等被统一裁剪,确保特征一致性。数据集以多种分辨率版本发布,包括原始RAW格式(保持最小边对齐至1400像素)、短边不超过800或1200像素的压缩版本,以及通过三阶段裁剪策略生成的480x800和480x1200像素高质量子集,满足不同训练需求。
特点
该数据集最显著的特点在于其精细化的多版本组织结构。它提供了五种不同规格的数据包,从保留元信息的RAW格式到经过缩放和裁剪的IMG+TXT配对数据,适配从高保真训练到高效推理的多元场景。特别设计的stage3版本通过三阶段裁剪技术,确保图像有效区域不低于480x480像素,同时将样本数从原始的336扩充至736,显著增加了数据多样性。此外,数据集附带了详尽的标签聚类分析,以文本和表格两种形式呈现,揭示了角色在不同服饰(如黑水手服、兔女郎装、学校泳装等)下的视觉模式,为风格迁移和角色定制提供了结构化参考。
使用方法
研究者可便捷地通过Hugging Face Hub下载该数据集。对于使用waifuc框架的用户,推荐加载RAW格式数据,通过`hf_hub_download`函数获取压缩包并解压至本地目录后,利用`LocalSource`接口即可轻松遍历图像及其元信息(如文件名、标签)。对于偏好直接图像-文本对训练的用户,可选择800、1200或stage3系列的ZIP包,解压后即可获得对齐的图像与标签文件。数据集采用MIT许可证开放,允许广泛的学术与商业应用,但需注意其内容可能不适宜所有受众,建议在敏感场景中审慎使用。
背景与挑战
背景概述
在文本到图像生成任务蓬勃发展的当下,高质量、细粒度的角色数据集成为推动模型性能跃升的关键基石。CyberHarem/mikazuki_kantaicollection 数据集由 DeepGHS 团队于近期构建,聚焦于《舰队Collection》中的角色“三日月”,旨在为动漫风格角色生成提供标准化训练资源。该数据集包含 336 张经过多源爬取(如 Danbooru、Pixiv、Zerochan)的高清图像,并附有详尽的标签体系,核心特征涵盖长发、黑发、呆毛、黄瞳及眉眼间距等关键视觉属性。通过提供多种分辨率版本(如 800px、1200px)及三级裁剪处理的数据包,该数据集显著降低了研究者获取规范化角色数据的门槛,成为动漫图像生成领域极具影响力的专项资源。
当前挑战
该数据集所应对的领域挑战集中于动漫角色生成的细粒度控制与数据稀缺性矛盾。一方面,现有通用数据集难以精准捕捉特定角色的标志性视觉元素(如三日月的水手服与月牙发饰),导致生成结果常出现特征混淆或遗漏;另一方面,从多源异构平台自动爬取图像时面临版权合规、图像质量参差及标签噪声等问题。在构建过程中,团队需处理跨平台元数据格式不一致、自动裁剪算法对构图破坏性风险,以及如何通过标签聚类(如识别出 6 个服饰簇)实现有效的数据增强。此外,在仅 336 张原始图像的基础上,如何通过多分辨率扩展与三级裁剪策略平衡数据规模与存储效率,亦是技术实现上的核心挑战。
常用场景
经典使用场景
该数据集以《舰队Collection》中的经典角色“三日月”为核心,汇聚了336张高质量图像及其精细标注的标签信息,为文本到图像生成任务提供了极具代表性的训练素材。其经典使用场景在于驱动基于扩散模型的二次元角色定制化生成,研究者可借助该数据集中包含的‘长发’、‘黑发’、‘呆毛’等核心特征标签,精准训练模型生成符合角色设定的肖像或全身图像。此外,数据集提供的多分辨率版本(如800像素与1200像素)以及三级裁剪版本,极大便利了不同计算资源下的模型微调与评估,成为风格迁移和角色一致性生成研究的理想基准。
衍生相关工作
该数据集衍生了一系列关于动漫角色聚类与多视角生成的前沿工作。基于其提供的标签聚类结果(如将图像按服饰或姿态分为6个簇),研究者开发了无监督的角色属性解析算法,能够自动挖掘未见角色的潜在风格模式。同时,数据集的原始版本与Waifuc库的集成催生了自动化数据采集与增强流程,后续工作进一步扩展至跨角色风格迁移和基于文本引导的局部编辑任务。这些成果不仅提升了二次元图像生成的效率,也为小样本角色学习领域贡献了可复现的评估框架。
数据集最近研究
最新研究方向
在二次元角色图像生成与个性化微调的前沿探索中,基于《舰队Collection》中三日月这一角色的专有数据集CyberHarem/mikazuki_kantaicollection正日益受到关注。该数据集汇集了336张高精度图像及其对应的特征标签,涵盖了从原始高清素材到多尺度裁剪版本等多种规格,为文本到图像生成模型提供了结构化的训练基础。当前研究方向聚焦于利用此类精细化、角色专属的数据集来优化扩散模型在动漫风格下的概念学习能力,尤其是在保持角色核心特征(如长发、黑发、呆毛与黄瞳)的同时,实现风格迁移与跨模态生成。随着Stable Diffusion等模型在二次元创作社区的广泛应用,该数据集不仅服务于角色拟真与服装变体的自动生成,还推动了少样本学习与标签聚类技术在虚拟角色个性化定制中的实际落地,成为连接游戏IP数字化与AI艺术创作的重要桥梁。
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务