20180117-2 刘子纬：Deep Learning Human-centric Representation in the Wild

报告嘉宾：刘子纬（加州伯克利）

报告时间：2018年01月17日（星期三）中午14:00（北京时间）

报告题目：Deep Learning Human-centric Representation in the Wild

主持人：王兴刚（华中科技大学）

报告摘要：

Understanding humans from photographs has been a long-pursuing goal of computer vision. Recently the advances of deep learning create lots of potentials in inducing effective representations even under complex scenarios. In this project, we comprehensively study automatic human-centric analysis with an emphasis on learning structural and hierarchical representations. Specifically, we unify the modeling of human-centric identity, attribute and landmarks in a coherent framework. Large-scale databases are collected and novel neural representations are proposed. Our approach has shown its effectiveness in the domain of face, fashion, parsing and motion understanding. Our findings have also been incorporated into several popular softwares and benchmarks.

报告相关文献列表：

[1] Z. Liu, P. Luo, X. Wang, X. Tang, “Deep Learning Face Attributes in the Wild”, ICCV, 2015.

[2] Z. Liu, P. Luo, S. Qiu, X. Wang, X. Tang, “DeepFashion: Powering Robust Clothes Recognition and Retrieval with Rich Annotations”, CVPR, 2016.

[3] Z. Liu, X. Li, P. Luo, C.C. Loy, X. Tang, “Deep Learning Markov Random Field for Semantic Segmentation”, TPAMI, 2017.

[4] Z. Liu, R. Yeh, X. Tang, Y. Liu, A. Agarwala, “Video Frame Synthesis by Deep Voxel Flow”, ICCV, 2017.

报告人简介：

Dr. Ziwei Liu is currently a post-doctoral research fellow at University of California, Berkeley. Prior to that, Ziwei received his PhD from the Chinese University of Hong Kong, under the supervision of Prof. Xiaoou Tang and Prof. Xiaogang Wang. He is also fortunate to have internships at Microsoft Research and Google Research. His research interests include computer vision, machine learning and computational photography.

讲者个人主页：

https://liuziwei7.github.io/

特别鸣谢本次Webinar主要组织者：

VOOC责任委员：王兴刚（华中科技大学）

VODB协调理事：王乃岩（北京图森未来科技有限公司）

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群已满，除讲者等嘉宾外，只能申请加入VALSE G群，群号：669280237），直播链接会在报告当天（每周三）在VALSE微信公众号和VALSE QQ群发布；

*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10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，请勿送花、打赏等，也不要说无关话语，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一推送上一周Webinar报告的总结及视频（经讲者允许后），每周四发布下一周Webinar报告的通知。

[slides]