时间：2011年10月20日上午10:00（周四）

地点：学术会议中心二楼报告厅

报告人：田奇 教授

单 位：美国德州大学圣安东尼奥分校

题 目：Coding for Large-scale Partial-duplicate Image Search

摘 要：Bag-of-visual-words model is widely used in the state-of-the-art large-scale image retrieval system. It represents each image as a bag of visual words by quantizing local image descriptors to the closest visual words. However, feature quantization reduces the discriminative power of local features, which causes many false visual word matches. Recently, some geometric verification methods are proposed to check the geometric consistency of matched features in a post-processing step. Although retrieval precision is improved, either the computational cost is too expensive to ensure real-time response, or they are limited to local verification. To address this dilemma, we propose a novel scheme, Spatial Coding and its variant Neighborhood Coding, designed for large scale partial-duplicate image retrieval. The spatial relationships among visual words are encoded in global region maps. Based on the region maps, a spatial verification approach is developed, which can detect false matches of local features efficiently, and consequently improve retrieval performance greatly.

Experiments in partial-duplicate image retrieval, using a database of one million images from Image-Net, reveal that our approach can effectively detect duplicate images with rotation, scale changes, occlusion, and background clutter with very low computational cost. The spatial coding and neighborhood coding achieve an 53% and 29.6% improvement in mean average precision and 46% and 67.6% reduction in time cost over the baseline Bag-of-Visual-Words approach, respectively. They perform even better than full geometric verification while being much less computationally expensive. Our demo on 10-million dataset further reveals the scalability of our approach.

报告人简介：田奇教授于1992年在清华大学电子工程系获得学士学位，2002在美国伊利诺伊州厄巴纳-香槟大学电子与计算机工程系获得博士学位。田奇教授现担任美国德州大学圣安东尼奥分校计算机科学系副教授、博士生导师。他还曾兼任微软亚洲研究院媒体计算组主任研究员以及资讯顾问、美国伊利诺伊大学访问学者、美国NEC研究院访问教授、美国三菱剑桥研究院访问研究员等职。田奇教授也是2010年国家自然科学基金委国际合作基金获得者。

田奇教授在多媒体领域进行了多年的研究有丰富的研究经验和重要的研究成果。其提出的创新学术思想推动了多媒体研究的发展，并已经被学术界大量引用。他在国际期刊/中文期刊和国际会议上已发表论文100余篇，其中40余篇发表在国际一流学术期刊如：IEEE TPAMI, IEEE TCSVT, IEEE Signal Processing Magazine, ACM Transactions on Knowledge Discovery from Data, IEEE Transactions on Computational Biology and Bioinformatics，IEEE Transactions on Multimedia, Pattern Recognition, IJCV等。90余篇发表在国际一流学术会议如ACM Multimedia，IEEE CVPR，ACM CIKM等。他的工作曾在2006年ICASSP会议上获得最佳论文奖，在2007年 PCM会议中获得最佳论文奖提名，其提出的判别式EM算法(DEM) 以及核判别EM算法(KDEM)已经被引用200余（Google Citations）次。田奇教授在国际学术界具有重要的影响力。他在很多国际期刊中担任重要职务，其中包括IEEE CSVT 副主编，IEEE TMM客座编辑， ACM TIST客座编辑和Elsevier CVIU杂志客座编辑等职务等。田奇教授还在很多顶级国际会议中担任主席和会议组织者，其中包括ACM Multimedia 2009 2010，ICPR 2010， CIVR2010，ICME 2010 等。