目前書報討論已改為線上課程，欲旁聽的「電機系」同學可以寄信給助教(chenwy0806@gapp.nthu.edu.tw)索取直播連結

 

Date&Time : 

2021 / 06 / 04 (Fri) 14:20

 

Location : 

Online

 

Speaker : 

 詹宗翰博士 (Dr. Tsung-Han Chan) 
聯發科技 (Multimedia Development, MediaTek)

 

Topic : 

Camera quality tuning: Challenges and opportunities

 

Abstract : 

Commodity imaging systems rely on power-efficient hardware image signal processing (ISP) pipelines. The configuration parameters of these ISPs often have complex interactions with the output color images, reconstructed from input RAW sensor measurements. In industry, the ISP parameters are manually tuned by “golden eye” experts or algorithm developers, prior to deployment according to application-specific quality metrics. The downside of such a manual process is that it is extremely time consuming; i.e., on the order of several man-months of work per project. In this talk, I will introduce recent advances in “automatic” camera quality tuning, or ISP parameters optimization. Particularly, two key components are discussed: First, what kind of optimization algorithms commonly used in ISP parameters tuning? Second, how to define quantitative image quality measure that replicates the behavior of human visual system? The focus of this talk aims to provide introduction and understanding of practical challenges of camera quality tuning faced in industry, with the hope that the messages to be delivered could bring out some interesting ideas in audiences’ researches and development.

 

Bio :

Tsung-Han Chan received the Ph.D. degree from the Institute of Communications Engineering, National Tsing Hua University, Hsinchu, Taiwan, in 2009. He is currently working as a Senior Technical Manager with MediaTek Inc., Hsinchu, Taiwan. Before joining MediaTek, he was a visiting Doctoral Graduate Research Assistant with Virginia Polytechnic Institute and State University, Arlington, in 2008, and a Postdoctoral Research Fellow with NTHU from 2009 to 2012, and a Research Scientist at Advanced Digital Sciences Center (a joint research center between University of Illinois - Urbana Champaign (UIUC) and A*STAR), Singapore in 2012 and 2013. He was also a Staff Engineer at Sunplus Technology Corp. in 2014.

In 2020, Dr. Chan was awarded MediaTek Special Contribution Award – Operation. He was also recognized as an Outstanding Reviewer for IEEE Computer Vision and Pattern Recognition 2014, and also recognized a Best Reviewer of the IEEE Transactions on Geoscience and Remote Sensing 2014. He was a co-recipient of the WHISPERS 2011 Best Paper Award.

His research interests are in image processing, numerical optimization and machine learning, with an emphasis on computer vision, image processing and deep learning for image quality assessment. In the research area, he has published more than 50 international papers with 3584 Google citations and 22 h-index.