主讲人简介：
卢宏涛，现为上海交通大学计算机系教授。研究方向为机器学习、模式识别、计算机视觉和复杂网络等等。在国内外学术期刊和顶级会议发表论文100多篇，发表的论文SCI他引1200多次，Google他引3000多次。连续入选2014、2015年Elsevier计算机科学中国高被引学者榜单。入选2005年教育部新世纪优秀人才计划，2003年上海市曙光学者。获两项省级自然科学和科技进步二等奖。主持863项目、国家自然基金项目（4项）等多项。
讲座内容简介：
Hashing is a technique for large-scale image retrieval, which represents images as compact binary codes so that the storage can be reduced and similarity computation can be significantly accelerated. In this talk, I’d like to present some of our works for image hashing in recent years which include the natural supervised hashing, the online self-organization hashing etc..
Deep learning has recently achieved great success in many fields such as speech recognition and computer vision, and has so attracted much attention from academic and industry communities. In this talk, I will present some of our recent researches on computer vision based on deep learning. The first is a method for click-through-rate (CTR) prediction in display advertising based on convolutional neural networks. The second is deep learning methods for large scale scene classification. I will also give some examples of deep learning based applications in computer vision such as activity recognition and face detection.
