Attending three doctoral research presentations in one day is overwhelming. Pil Ho Kim and Bin Liu had their thesis proposal presentation to their committee and Matt Wolenetz (Kishore Ramachandranâ€™s student) had his doctoral defense today at Georgia Tech. Before these three I had a meeting with Nikil Jayant and Jeannie Lee to discuss her research directions. So I had an interesting day in Atlanta. All of these are obviously different topics, but related. Bin Liu started with his work on continuous queries on Live Multimedia data. Pil Ho presented his unified indexing and tagging framework for eChronicles, and Matt defended his work on energy optimization in sensor networks.
One common problem that most doctoral students, and even many senior researchers, face is their desire to present details without presenting the big picture and where their research fits. This really hurts their presentation because people donâ€™t understand why their research is important. It is clear that many researchers believe that they should be presenting details of their work. In doing that they always get lost in detailed formalisms and jargons that makes their presentations understandable only by a very few people. Sometimes I feel that many of my colleagues take satisfaction in thinking that their work is so clever that only a few people can understand that.
It was nice to see that my two remaining students at Georgia Tech have successfully crossed this phase of their education. Now they have to focus on finishing remaining part of their research and write their dissertation. Querying live multimedia data is a very important problem that will become increasingly important in many applications, particularly on the Web. Currently there are no techniques in this area. Bin can make a very important contribution in this space. Lately many research groups in many different countries have started considering eChronicles (with different names, however) a very important topic. Unified indexing of spatio-temporal data is a key issue in organizing eChronicles and indexing them for proper use. Pil Ho is making good progress in this area. Gopal Pingali at IBM is championing this area in IBM for enterprise applications.
I am eagerly waiting to see how far Bin Liu and Pil Ho can go in their work and how to build these areas to develop experiential computing further.
Thanks for your kind words.
I’d be happy to share my opinions with you. There are several places that have good programs. But most academic places do not really care much beyond lip service about productization or real applications. In the context of Image and Video search I will be giving a talk in Singapore on next friday. I will share my presentation here in my blog.
Feel free to send me mail directly for any specific questions at:
A very interesting review..