设为首页收藏本站

VALSE

VALSE 首页 活动通知 查看内容

20190327-06 人脸识别:路在何方?FaceRecognition:WheretoGo?

2019-3-25 15:07| 发布者: 程一-计算所| 查看: 1340| 评论: 0

摘要: VALSE Webinar改版说明:自2019年1月起,VALSE Webinar改革活动形式,由过去每次一个讲者的方式改为两种可能的形式:1)Webinar专题研讨:每次活动有一个研讨主题,先邀请两位主题相关的优秀讲者做专题报告(每人30 ...

VALSE Webinar改版说明:

自2019年1月起,VALSE Webinar改革活动形式,由过去每次一个讲者的方式改为两种可能的形式:

1)Webinar专题研讨:每次活动有一个研讨主题,先邀请两位主题相关的优秀讲者做专题报告(每人30分钟),随后邀请额外的2~3位嘉宾共同就研讨主题进行讨论(30分钟)。

2)Webinar特邀报告:每次活动邀请一位资深专家主讲,就其在自己熟悉领域的科研工作进行系统深入的介绍,报告时间50分钟,主持人与主讲人互动10分钟,自由问答10分钟。


报告时间:2019年3月27日(星期三)晚上21:00(北京时间)

主题:人脸识别:路在何方?Face Recognition: Where to Go?

主持人:山世光(中国科学院计算技术研究所


报告嘉宾:刘小明(Michigan State University)

报告题目:3D Face Modeling, Reconstruction and its Role in Face Recognition


报告嘉宾:华刚(Wormpex AI Research)

报告题目:Efficient, Accurate, and Robust Face Recognition


Panel议题:

  1. 人脸识别是一个solved problem吗?现有SOTA系统还有什么不足需要继续深挖?学术界、工业界和用户对识别系统性能的评价有何不同理解?何时可以说一个问题是solved?

  2. 学术界应该如何面对工业界靠数据取胜的局面?除了靠数据,还有什么可以提高性能?比如现在很多系统对儿童的识别会差一些,是数据问题还是算法问题?

  3. 学术界需要什么样的新benchmark?人脸识别中的小样本问题如何定义?Open-set问题又该如何定义?

  4. 1:1场景下,验证率不低于95%时,现有SOTA人脸识别系统的FAR能到亿分之一吗?

  5. 在1:N场景,N=10万人,且Open-set场景下,在识别率不低于95%时,FAR能到0.1%或更低吗?

  6. 现有人脸识别系统什么情况下比人类强?什么时候不如?

  7. 是否有必要以及如何提升现有人脸识别系统的可解释性?


Panel嘉宾:

乔宇(中国科学院深圳先进技术研究院)杨恒(深圳爱莫科技有限公司)


*欢迎大家在下方留言提出主题相关问题,主持人和panel嘉宾会从中选择若干热度高的问题加入panel议题!

报告嘉宾:刘小明(Michigan State University)

报告时间:2019年3月27日(星期三)晚上21:00(北京时间)

报告题目:3D Face Modeling, Reconstruction and its Role in Face Recognition


报告人简介:

Xiaoming Liu is an Associate Professor at the Department of Computer Science and Engineering of Michigan State University. He received the Ph.D. degree in Electrical and Computer Engineering from Carnegie Mellon University in 2004. Before joining MSU in Fall 2012, he was a research scientist at General Electric (GE) Global Research. His research interests include computer vision, patter recognition, biometrics and machine learning. He is the recipient of 2018 Withrow Distinguished Scholar Award from Michigan State University. As a co-author, he is a recipient of Best Industry Related Paper Award runner-up at ICPR 2014, Best Student Paper Award at WACV 2012 and 2014, and Best Poster Award at BMVC 2015. He has been an Area Chair for numerous conferences, including FG, ICPR, WACV, ICIP, ICCV, and CVPR. He is the Co-Program Chair of BTAS 2018 and WACV 2018 conference. He is an Associate Editor of Neurocomputing journal and Pattern Recognition Letters. He is a guest editor for International Journal of Computer Vision (IJCV) Special Issue on Deep Learning for Face Analysis, and ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM) Special Issue on Face Analysis for Applications. He has authored more than 100 scientific publications, and has filed 26 U.S. patents. His work has been cited over 5000 times according to Google Scholar, and has an H-index of 40.


个人主页:

http://cvlab.cse.msu.edu


报告摘要:

After many years of face recognition research, one question in the community is how to push face recognition to the next level. We view that utilizing 3D geometry information in the recognition pipeline is one potential avenue to grow. This talk will present some of our recent work in 3D face modeling from a collection of in-the-wild images, or raw scans. We will also discuss how the increasingly accurate 3D geometry information may benefit face recognition systems.


参考文献:

[1]  Luan Tran, Xiaoming Liu, “Nonlinear 3D Face Morphable Model,” in Proceeding of IEEE Computer Vision and Pattern Recognition (CVPR 2018), Salt Lake City, UT, June 2018.

[2]  Feng Liu, Luan Tran, and Xiaoming Liu, "3D Face Modeling from Diverse Raw Scan Data," arXiv preprint arXiv:1902.04943 (2019).

报告嘉宾:华刚(Wormpex AI Research)

报告时间:2019年3月27日(星期三)晚上21:30(北京时间)

报告题目:Efficient, Accurate, and Robust Face Recognition


报告人简介:

Gang Hua is currently Vice President and Chief Scientist of Wormpex AI Research (便利蜂人工智能研究院). Before that, he served at various roles at Microsoft including Technical/Science Advisor of CVP, Director of Computer Vision Science in Redmond and Taipei ATL, and Principal Researcher/Research Manager. His research focuses on computer vision, pattern recognition, machine learning, robotics, towards general Artificial Intelligence. He is the recipient of the 2015 IAPR Young Biometrics Investigator Award for his contribution to Unconstrained Face Recognition from Images and Videos. He is a Program Chair for CVPR'2019 and CVPR’2022. He has served as Area Chairs for many top international conferences. He is currently an Associate Editor in Chief for CVIU, and Associate Editors for IJCV, IEEE T-IP, IEEE T-CSVT, IEEE Multimedia, and MVA. He is a IEEE Fellow, an IAPR Fellow, and an ACM Distinguished Scientist.


个人主页:

http://www.ganghua.org


报告摘要:

Face recognition has matured to a stage to support a lot of commercial applications. Nevertheless, there are still lots of challenges need to be addressed. I will start the talk with a brief status quo of face recognition research and development in both academia and industry. Then, I will summarize some of our recent research works on efficient and accurate face detection and recognition, and how we may deal with adversarial attacks using an identity preserving deep generative models. I will conclude my talks with recent trends in the research community in face recognition.


参考文献:

[1] Jianmin Bao*, Dong Chen, Fang Wen, Houqiang Li, and Gang Hua, "Towards Open-Set Identity Preserving Face Synthesis," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR'2018), Salt Lake City, Utah, June 2018.

[2] Jianmin Bao*, Dong Chen, Fang Wen, Houqiang Li, and Gang Hua, "CVAE-GAN: Fine-Grained Image Generation through Asymetric Training," in Proc. IEEE International Conf. on Computer Vision (ICCV'2017), Venice, Italy, October, 2017. 

[3] Jiaolong Yang*, Peiran Ren, Dongqing Zhang, Dong Chen, Fang Wen, Hongdong Li, and Gang Hua, "Neural Aggregation Networks for Video Face Recognition," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR'2017), Honolulu, Hawaii, July 2017.

Panel嘉宾:乔宇(中国科学院深圳先进技术研究院)

嘉宾简介:

中国科学院深圳先进技术研究院研究员,数字所所长,广东省机器视觉与虚拟现实重点实验室常务副主任。入选科技部中青年科技创新领军人才、是广东省特支计划科技创新领军人才、深圳鹏程学者长期特聘教授。从事计算机视觉、深度学习和机器人等领域的研究,已在包括PAMI,T-IP,IJCV,CVPR, ICCV,ECCV,AAAI等重要国际会议和期刊上发表学术论文170余篇,多次在ChaLearn, LSun, ActivityNet,EmotionW等国际评测中取得第一,获ImageNet 2016场景分类任务第二名。


个人主页:


http://mmlab.siat.ac.cn/yuqiao/

Panel嘉宾:杨恒(深圳爱莫科技有限公司)

嘉宾简介:

Heng Yang, Founder and CEO of Shenzhen AiMall Tech since 2018, before that he worked as the CTO of a startup. He received  his PhD degree from Queen Mary University of London with thesis on Face Alignment in the Wild in 2015. Before joining industry in 2016, he worked as a research associate at University of Cambridge on the topic of animal health monitoring by looking at the face. He has published around 20 academic papers including CVPR/ICCV/TIP. He has rich experience in face application in industry and provided enormous solutions  (face tracking, recognition and liveness detection, etc)  to many international clients.


个人主页:

https://www.linkedin.com/in/heng-yang-88249946/


19-06期VALSE在线学术报告参与方式:


长按或扫描下方二维码,关注“VALSE”微信公众号(valse_wechat),后台回复“06期”,获取直播地址。



特别鸣谢本次Webinar主要组织者:

主办AC:欧阳万里(The University of Sydney)

责任AC:王楠楠(西安电子科技大学)、杨猛(中山大学)、杨恒(深圳爱莫科技有限公司)


活动参与方式:

1、VALSE Webinar活动依托在线直播平台进行,活动时讲者会上传PPT或共享屏幕,听众可以看到Slides,听到讲者的语音,并通过聊天功能与讲者交互;

2、为参加活动,请关注VALSE微信公众号:valse_wechat 或加入VALSE QQ群(目前A、B、C、D、E、F、G、H、I群已满,除讲者等嘉宾外,只能申请加入VALSE J群,群号:734872379);

*注:申请加入VALSE QQ群时需验证姓名、单位和身份,缺一不可。入群后,请实名,姓名身份单位。身份:学校及科研单位人员T;企业研发I;博士D;硕士M。

3、在活动开始前5分钟左右,讲者会开启直播,听众点击直播链接即可参加活动,支持安装Windows系统的电脑、MAC电脑、手机等设备;

4、活动过程中,请不要说无关话语,以免影响活动正常进行;

5、活动过程中,如出现听不到或看不到视频等问题,建议退出再重新进入,一般都能解决问题;

6、建议务必在速度较快的网络上参加活动,优先采用有线网络连接;

7、VALSE微信公众号会在每周四发布下一周Webinar报告的通知及直播链接。

8、Webinar报告的PPT(经讲者允许后),会在VALSE官网每期报告通知的最下方更新[slides]

9、Webinar报告的视频(经讲者允许后),会更新在VALSE爱奇艺空间,请在爱奇艺关注Valse Webinar进行观看。


刘小明[slides]

华刚[slides]

Archiver|手机版|小黑屋|Vision And Learning SEminar    

GMT+8, 2019-6-24 23:44 , Processed in 0.041046 second(s), 16 queries .

Powered by Discuz! X3.2

© 2001-2013 Comsenz Inc.

返回顶部