报告题目:Recent Advances in Estimating Air Quality from Single Photo of Daily Living
报告时间:2017年1月19号(周四)上午10点
报告地点:信息学院四楼报告厅
报告人:孟雷博士
报告人单位:南洋理工大学,新加坡
报告人简介:
孟雷博士现任研究员,就职于新加坡南洋理工大学百合卓越联合研究中心。目前研究重点聚焦在面向电子商务和老年健康的机器学习与大数据分析问题。担任“银发丝绸之路(Silver Silk Road)”项目组负责人,负责在线多模态的电商产品索引和检索系统研发。同时,在“银发延展(Silver Extension)”项目组中承担支持非打扰式居家养老系统在线数据分析算法的研究。
研究方向:机器学习、数据挖掘、社交媒体分析,主要开展聚类算法、异构信息融合方法、多模态信息索引和检索算法和模式识别算法的理论研究及其在大规模社交媒体和社交网络等领域的应用。
在国内外著名期刊和会议TKDE、 TNNLS、 SDM、 ICMR以及 IJCNN上发表学术论文十余篇,担任Applied Soft Computing期刊的编委,并担任KDD、ICDM、SDM、TNNLS以及Applied Intelligence等会议和期刊的审稿人,是国际神经网络协会(INNS)和电子与电气工程师协会(IEEE)的会员。
报告摘要:
The issue of particulate matter in the air is a global question, challenging the health of people, especially children and the aging population. Thus, knowing the quality of air around them becomes an important step for daily activity arrangement. Besides the way of purchasing sensors, an emerging trend in AI research is to explore the use of phone photos for air quality estimation, by leveraging the rich haze photos available on the Web and advanced machine learning techniques.
This talk will first elaborate recent advances in air quality prediction from single photo of daily living, including technical challenges and the state-of-the-arts. Thereafter, a briefing of the burning AI technology, namely, the deep learning techniques, will be given. Lastly, our newly proposed solution, based on fast haze region segmentation and deep learning, will be presented.