哥伦比亚大学公众人物脸部数据库

IntroductionThe PubFig database is a large, real-world face dataset consisting of58,797images of200people collected from the internet. Unlike most other existing face datasets, these images are taken in completely uncontrolled situations with non-cooperative subjects. Thus, there is large variation in pose, lighting, expression, scene, camera, imaging conditions and parameters, etc. The PubFig dataset is similar in spirit to theLabeled Faces in the Wild (LFW) datasetcreated at UMass-Amherst, although there are some significant differences in the two:LFW contains13,233 imagesof5,749 people, and is thus much broader than PubFig. However, it's also smaller and much shallower (many fewer images per person on average).LFW is derived from theNames and Faces in the Newswork of T. Berg, et al. These images were originally collected using news sources online. For many people, there are often several images taken at the same event, with the person wearing similar clothing and in the same environment.Our paper at ICCV 2009showed that this can often be exploited by algorithms to give unrealistics boosts in performance.Of course, the PubFig dataset no doubt has biases of its own, and we welcome any attempts to categorize these.CitationThe database is made available only for non-commercial use. If you use this dataset, please cite the following paper:"Attribute and Simile Classifiers for Face Verification,"Neeraj Kumar, Alexander C. Berg, Peter N. Belhumeur, and Shree K. Nayar,International Conference on Computer Vision (ICCV), 2009.[bibtex] [pdf] [webpage]Related ProjectsAttribute and Simile Classifiers for Face Verification(Columbia)FaceTracer: A Search Engine for Large Collections of Images with Faces(Columbia)Labeled Faces in the Wild(UMass-Amherst)Names and Faces(SUNY-Stonybrook)

引言
PubFig数据库是一款大型真实世界人脸数据集，包含从互联网采集的200位人物共计58797张图像。与绝大多数现有公开人脸数据集不同，本数据集的所有图像均拍摄于完全非受控场景，且被摄对象均处于非配合状态。因此，图像在姿态、光照、表情、场景、相机参数、成像条件等诸多维度均存在极大差异。

PubFig数据集的设计理念与马萨诸塞大学阿默斯特分校构建的野外标注人脸（Labeled Faces in the Wild，LFW）数据集相近，但二者存在显著区别：LFW数据集涵盖5749位人物的13233张图像，覆盖人群范围比PubFig更广，但整体数据规模更小，且单人均值图像数更少（平均每人拥有的图像数量远低于PubFig）。

LFW数据集源自T. Berg等人的《新闻中的姓名与人脸》研究项目，其原始图像通过在线新闻资源采集得到。对于多数人物而言，数据集内常包含同一事件下拍摄的多张图像，人物着装相似且拍摄环境一致。我们在2009年国际计算机视觉会议（International Conference on Computer Vision，ICCV）上发表的论文指出，算法常可利用这一数据特性获得不切实际的性能提升。

当然，PubFig数据集本身无疑存在固有偏差，我们欢迎各界对其偏差展开分类研究。

# 引用说明
本数据库仅可用于非商业用途。若您使用该数据集，请引用以下论文：
《用于人脸验证的属性与相似性分类器》（*Attribute and Simile Classifiers for Face Verification*），作者为Neeraj Kumar、Alexander C. Berg、Peter N. Belhumeur及Shree K. Nayar，发表于2009年国际计算机视觉会议（ICCV）。
[bibtex] [pdf] [webpage]

# 相关项目
1. 用于人脸验证的属性与相似性分类器（哥伦比亚大学）
2. FaceTracer：面向大规模人脸图像集的搜索引擎（哥伦比亚大学）
3. 野外标注人脸（Labeled Faces in the Wild，UMass-Amherst）
4. 姓名与人脸（Names and Faces，纽约州立大学石溪分校）

搜集汇总

背景与挑战

背景概述

哥伦比亚大学公众人物脸部数据库（PubFig）是一个大规模、真实世界的人脸数据集，包含58,797张图像，涵盖200位公众人物，所有图像均从互联网采集，在非受控环境下拍摄，主体非合作，因此具有姿态、光照、表情和场景等方面的巨大变化。该数据集专为研究设计，仅限非商业使用，常用于人脸验证等计算机视觉任务，与LFW数据集相比，它更深入但范围更窄。