新闻公告
【7月19日】山东大学聂礼强教授学术报告

    报告题目:面向微视频的多模态分析技术

    报告人:聂礼强 教授

    时间:7月19日下午17:00-18:00

    地点:信息学院北楼A404


    报告摘要:

    Bite-sized videos enforcing the shorter-is-better strategy, are taking the social media world by storm with the rising of the micro-video sharing services, like Vine, Snapchat, and Viddy. Micro-videos can benefit lots of commercial applications, such as brand building. Despite their value, the analysis and modeling of micro-videos is non-trivial due to the following reasons: 1) micro-videos are short in length and of low quality; 2) they can be described by multiple heterogeneous channels, spanning from social, visual, and acoustic to textual modalities; and 3) there are no available benchmark dataset. In this research, we build a large-scale dataset and attempt to solve two sub-problems of micro-videos: venue category estimation and popularity prediction. We jointly learn the optimal latent common space from multi-modalities, whereby the popularity or venue category of micro-videos can be better identified. From the common latent spaces found, we then develop a tree-guided multi-task multi-modal learning model to estimate the venue category, and a novel transductive multi-modal learning method to predict the popularity of micro-videos. We validated the effectiveness of the models on our collected dataset. The data and codes have been released to facilitate other researchers. This talk presents our current research and future work towards leveraging micro-videos for better social media analytics.


    报告人简介:

    Dr. Liqiang Nie is currently a professor with Shandong University. He respectively received his B.E. degree from Xi’an Jiaotong University in 2009, and the Ph.D. degree from National University of Singapore (NUS) in 2013. After Ph.D., he worked 3 plus years in NUS as a research fellow. His research interests include multimedia computing, information retrieval, and their applications in wellness computation. He has published around 100 technical papers in the international conferences and journals, such as SIGIR, ACM MM, TOIS, and TKDE. Thereinto, around 60 papers are at the venues of CCF A or IEEE/ACM Trans. Dr. Nie has served as guest editors of some reputable journals like IEEE Trans Big Data, ToMM, and special session organizers of some conferences, such as ICMR and MMM. He is an associate editor of information science and a program chair of ICIMCS 2017.