Tao-Chun Lee

Audiovisual Communications Laboratory (LCAV )
School of Computer and Communication Sciences (IC)
Ecole Polytechnique Fédérale de Lausanne (EPFL)
Office: BC 325
Phone: +41 (0)21 693 8158
Fax: +41 (0)21 693 4312
Email: tao.lee [at] epfl.ch
Address: EPFL – I&C – LCAV
BC 325 (Building BC)
Station 14
1015 Lausanne, Switzerland



 With the the wall painting of Jungfrau as the background, I introduce myself as a man passionate about efficient audiovisual presentation of information. I received my alma mater in electrical engineering from National Tsing-Hua University in Taiwan, and my engineer diploma in informatics from EPFL. I am interested in online education of informatics, in particular, signal processing, programming methods, and computer systems. I believe the advance of online education can bring informatics closer to the young generation. I enjoy Toastmasters, Iaido practice, and building websites in my free time. Check out the website of Taiwanese Student Association.  



  •  Doctoral school, EPFL, 10.2013 ~ 10.2014
  •  Engineer diploma in informatics, EPFL, 09.2013



 I am generally interested in the analysis and design of Massive Open Online Courses (MOOCs), an emergent paradigm of large-scale knowledge distribution on the internet. The challenge we are facing is how to analyze and design the underlying structure of knowledge distribution. To address this challenge, we need good answers to the following questions.

  • How to model MOOCs?
  • How to understand the behavior of the participants in MOOCs?
  • How to estimate the learning capacity of MOOCs?
  • How to estimate the teaching capacity of MOOCs?
  • How to improve MOOCs?

Without overly generalizing the problem, we focus on crafting a pilot MOOC, the Digital Signal Processing (DSP) MOOC on Coursera, as an experimental playground to test the theory and the practice of MOOC research.

DSP MOOC on Coursera


My research business and my teaching business are intermately related to each other: My research is also my teaching. The teaching of an online class with more than 100,000 learners to date presents itself with unprecedented challenges. A MOOC provides great opportunities to observe the interaction between students, teaching material and teaching staffs. In particular,

  • How students struggle to master the material?
  • What kind of help students would expect?
  • How students ask questions and help each other?
  • Would numerical examples and online labs help students learn?
  • What topics are confusing to students?



 The rationals behind dropping out a PHD are sometimes even more passionate than applying for one. Here are mine:

  • When you know you will NOT be happy on the day you hand-in your PHD thesis
  • When your PHD thesis topic is a life-long pursuit rather than a four-year effort
  • When you do not want to pursue a research career afterwards
  • When you feel the level of research and teaching is at a standard beyond your reach
  • When you see good opportunities that you feel better like pursuing



T.-C. Lee; P. Prandoni; M. Vetterli : Analysis and Design of Large-scale Online Classes. 2014.
T.-C. Lee : Integrating Defect Data, Code Review Data, and Version Control Data for Defect Analysis and Prediction ; 2013.
T.-C. Lee; J. Unnikrishnan : MONITORING NETWORK STRUCTURE AND CONTENT QUALITY OF SIGNAL PROCESSING ARTICLES ON WIKIPEDIA. 2013. International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2013, Vancouver, British Columbia, Canada, May 26-31, 2013.
T.-C. Lee; M.-A. Boéchat : When the ARM weakly consistent memory model meets speculation: is it necessary? ; 2012.
T.-C. Lee; B. Fasel : Implementing a Cocktail-party Processor via Time-frequency Masking ; 2012.
T.-C. Lee; S. Pellone : Does Stanford's induction apply to engineering sciences? ; 2012.
T.-C. Lee; P. Prandoni : Reverse Digital Typography ; 2012.