Prof. Lin Li
Title: Multimodal Representation Learning
Professor Lin Li received her PhD from The University of Tokyo in 2009. She is currently a full professor at Wuhan University of Technology, China. She was a visiting scholar in Advanced Analytics Institute, University of Technology, Sydney, during 2014-2015. Her current research interest includes information retrieval, recommendation system and Web data mining. She has published over 60 research papers, including over 10 journal papers and papers at top-tier AI and data mining conferences such as AAAI, ICDM, ICMR, ICASSP, PAKDD, etc. She serves as the Editorial Board Member for Web Intelligence Journal and Human-Centric Intelligent Systems Journal, Guest Editor for World Wide Web Journal and a reviewer for ACM TOMM, ACM TOIS, ACM TIST, IEEE TKDE, WWWJ, KBS, etc.
With rapid development of social networks and search engines, a surge of interests has been witnessed in jointly analyzing of multimodal data such as text, image, audio and video. To cope with this situation, information acquisition and processing have to be transformed from the form of single media to multimedia. Therefore, challenges stemming from the “media gap”, which means that representations of different media types are inconsistent, are appealing increasing attention. Recently, deep neural networks(DNN), a major breakthrough in machine learning, have been employed to learn better multimodal representations. This talk introduces a taxonomy of multimodal machine learning from representation, translation, fusion, alignment, and co-learning. Our recent studies in multimodal representation are presented with new multimodal algorithms and exciting multimodal applications.