行健讲坛学术讲座

时间:  2018年12月17日（周一）上午10：00

地点:  校本部东区翔英楼T706室

讲座：基于密集网络的图像超分辨率

演讲者：A/ Prof. Jian Zhang, 澳大利亚悉尼科技大学（UTS）

演讲者简介：Dr. Jian Zhang received the BSc. degree from East China Normal University, Shanghai, China, in 1982; the MSc. degree in computer science from Flinders University, Adelaide, Australia, in 1994; and the Ph.D. degree in electrical engineering from the University of New South Wales (UNSW), Sydney, Australia, in 1999.

From 1997 to 2003, he was with the Visual Information Processing Laboratory, Motorola Labs, Sydney, as a Senior Research Engineer, and later became a Principal Research Engineer and a Foundation Manager with the Visual Communications Research Team. From 2004 to July 2011, he was a Principal Researcher and a Project Leader with National ICT Australia, Sydney. He is currently an Associate Professor with the Global Big Data Technologies Centre, School of Electrical & Data Engineering, Faculty of engineering and Information Technology, University of Technology Sydney, Sydney. Prof Zhang’s research interests include multimedia signal processing, computer vision, pattern recognition, visual information mining, human-computer interaction and intelligent video surveillance systems. Prof Zhang has co-authored more than 130 paper publications, book chapters, patents and technical reports from his research output, he was the co-author of eight granted US and China patents.

Dr. Zhang is an IEEE Senior Member. He was Technical Program Chair, 2008 IEEE Multimedia Signal Processing Workshop; Associated Editor, IEEE Transactions on Multimedia; Associated Editor, IEEE Transactions on Circuits and Systems for Video Technology (T-CSVT); Associated Editor, EURASIP Journal on Image and Video Processing. As a General Co-Chair, Jian chaired the International Conference on Multimedia and Expo (ICME 2012) in Melbourne Australia 2012. As a Technical Program Co-Chair, Jian chaired The IEEE Visual Communications and Image Processing (IEEE VCIP 2014).

讲座摘要：图像超分辨率是一类增加图像分辨率的技术，被广泛应用于需要高清图像的图像处理技术中。最近，深度学习方法被证明能有效处理图像超分辨率问题。其中，深密集网络在图像超分辨率上取得了很好的效果，这得益于密集层之间特征重用。但是，普通密集网络限制了块之间的特征重用。因此我们提出了用于图像超分辨率的双密集网络来提升特征学习性能。这个双密集网络通过加入了我们所提出块间密集连接，来扩展原本的块内的密集连接。这样，特征信息不但只传播到随后的一个块内，而是传给后来的所有块。因此，在深度网络训练过程中的梯度和特征消失的问题可以被解决。另外，我们发现训练用于图像超分辨率的普通密集网络，要消耗大量内存。为了减少训练时的内存占用，构建更深的网络， 我们引入共享内存分配的方法，来构建适用于图像超分辨率的内存优化的密集网络。在公开的基准的数据集上的评测结果表现，我们所提出的模型在使用适中的参数量和计算量的情况下，超过目前最新的超分辨率方法的图像重建效果。

邀请人：上海大学通信与信息工程学院 安平 教授

欢迎广大教师和学生参加！