报告摘要: Deep CNN-based object detection systems have achieved remarkable success on several large-scale object detection benchmarks. However, training such detectors requires a large number of labeled bounding boxes, which are more difficult to obtain than image-level annotations. Previous work addresses this issue by transforming image-level classifiers into object detectors. This is done by modeling the differences between the two on categories with both image-level and bounding box annotations, and transferring this information to convert classifiers to detectors for categories without bounding box annotations. We improve this previous work by incorporating knowledge about object similarities from visual and semantic domains during the transfer process. The intuition behind our proposed method is that visually and semantically similar categories should exhibit more common transferable properties than dissimilar categories, e.g. a better detector would result by transforming the differences between a dog classifier and a dog detector onto the cat class, than would by transforming from the violin class. Experimental results on the challenging ILSVRC2013 detection dataset demonstrate that each of our proposed object similarity based knowledge transfer methods outperforms the baseline methods. We found strong evidence that visual similarity and semantic relatedness are complementary for the task, and when combined notably improve detection, achieving state-of-the-art detection performance in a semi-supervised setting.

报告人简介：Prof. Liming Chen was awarded the joint B.Sc. degree in mathematics and computer science from the University of Nantes, Nantes, France, in 1984. He obtained the M.S. and Ph.D. degrees in computer science from the University of Pierre and Marie Curie Paris 6, Paris, France, in 1986 and 1989, respectively. He first served as an Associate Professor at the Université de Technologie de Compiègne, Compiègne, France, and then joined Ecole Centrale de Lyon, Ecully, France, as Professor in 1998, where he leads an advanced research team on multimedia computing and pattern recognition. From 2001 to 2003, he also served as Chief Scientific Officer in a Paris-based company, Avivias, specialized in media asset management. He has been Head of the department of Mathematics and Computer science from 2007. Prof. Liming Chen has taken out 3 patents, authored more than 200 publications and acted as chairman, PC member and reviewer in a number of high profile journal and conferences since 1995. He has been a (co)-principal investigator on a number of research grants from EU FP programme, French research funding bodies and local government departments. His current research interest is computer vision and multimedia, spanning from 2-D/3-D face analysis and recognition, image and video analysis and categorization to affect analysis both in image audio and video. Prof. Liming Chen holds the premium of scientific excellence attributed by the French Ministry of Higher Education and Research since 1995.

Liming Chen教授中文简介：Liming Chen教授于1984年获得法国南特大学(University of Nantes)的数学和计算机专业双学士学位。并分别与1986年和1989年获得法国皮埃尔玛丽居里巴黎六大(University of Pierre and Marie Curie Paris 6)计算机专业硕士和博士学位。之后，他首先以助理教授的身份任职于贡比涅技术大学 (Université de Technologie de Compiègne)。1995年，他获得法国高等教育研究部颁发的卓越科学贡献奖。1998年他以正教授身份加入里昂中央理工大学(Ecole Centrale de Lyon)，率先组建了多媒体计算和模式识别研究小组。经过多年积累，该小组在计算机视觉和多媒体处理，特别是二维和三维人脸分析和识别，基于语音、图像和视频的模式分类，情感分析等方面做出了重要学术成果。其研究获得了多项专利，发表了200多篇学术论文和专著，获得了多项欧盟和法国研究署的基金资助。此外，Chen教授担任多个高水平期刊的审稿人和会议程序委员。2001至2003年，他兼任巴黎Avivias公司的首席科学家，从2007年起，他担任数学与计算机学院院长。

欢迎感兴趣的师生参加！