20161012-34 李慧斌：Three-Dimensional Facial Identity and Expression Analysis... ...

报告嘉宾2: 李慧斌 （西安交通大学数学与统计学院信息科学系）

报告时间：2016年10月12日（星期三）晚21：00（北京时间）

报告题目：Three-Dimensional Facial Identity and Expression Analysis: From Handcrafted Features to Learned Features

主持人：郭裕兰（国防科技大学）

报告摘要：

In this talk, I will introduce the topic of 3D facial analysis, including 3D face recognition and 3D facial expression recognition. For 3D face recognition, I will introduce one of our previous work published in the journal of IJCV, which is a totally registration-free 3D face recognition approach based on the idea of matching mesh-based 3D local surface descriptors. To our knowledge, our approach achieved the best identification rate of 96.56% on the whole Bosphorus 3D face database before our submission. The source code of this work is publicly available at my homepage http://gr.xjtu.edu.cn/web/huibinli/code. For 3D facial expression recognition, I will introduce one of our ongoing work, which is a new attempt to use deep convolutional neutral network for 2D-3D multimodal facial expression recognition. Experimental results demonstrated that our method can achieve the best performance than handcrafted features, pre-trained deep models, fine-tuned deep models and all the state-of-the-art methods on the BU-3DFE database and Bosphorus database.

参考文献：

1. Huibin Li, Di Huang, Jean-Marie Morvan, Yunhong Wang, Liming Chen, Towards 3D Face Recognition, A Registration-Free Approach using Fine-Grained Matching of 3D Keypoint Descriptors, International Journal of Computer Vision (IJCV), 113(2): 128-142, 2015.

2. Huibin Li, Jian Sun, Dong Wang, Zongben Xu, Liming Chen, Deep Representation of Facial Geometric and Photometric Attributes for Automatic 3D Facial Expression Recognition, http://arxiv.org/abs/1511.03015, 2015.

3. Huibin Li, Jian Sun, Zongben Xu, Liming Chen, Multimodal 2D+3D Facial Expression Recognition with Deep Fusion Convolutional Neural Network, Submitted to IEEE Transaction on Multimedia, under major revision 2016.

报告人简介:

李慧斌，2006年本科毕业于陕西师范大学数学与应用数学专业，2009年硕士毕业于西安交通大学数学与信息科学专业，2013年获得法国里昂中央理工大学数学与计算机博士学位。2013年底加入西安交通大学数学与统计学院，被聘为新讲师。

主要从事三维人脸识别、三维面部表情识别、离散微分几何和几何逼近论等领域的交叉研究。其研究成果曾获第七届欧盟生物特征识别工业与学术奖的提名奖。现已发表期刊和会议论文19篇。现主持国家自然科学基金青年基金和中国博士后基金面上项目各一项。其代表性研究成果发表在IJCV、IEEE-TVCG、CVIU等期刊上，总引用次数170余次（google scholar）。现担任多个国际期刊（如IEEE-TIP、IEEE-TIFS、IEEE-TSMC Part C）和会议（如ICIP、ICPR、ICB、BTAS、NIPS）的审稿人。

特别鸣谢本次Webinar主要组织者：

VOOC责任委员：郭裕兰（国防科技大学）

VODB协调理事：张利军（南京大学），章国锋（浙江大学）

[Slides]