|
Introduction |
|
Hi, my name is Tanasai Sucontphunt, pronounced like Ta-na-sai Soo-kon-pan, (ธนาสัย สุคนธ์พันธุ์ in Thai). Welcome to my website!
I am a PhD student in the Department of Computer Science at University of Southern California. My main research is on Computer Graphics and its Application.
Here is my little space combining my research blog, publication, biography, portfolio, and review comments on almost everything.
I am a livelong learner. I am always excited if I found new technology emerging with my new idea to develop something useful.
Thanks for stopping by. |
|

|
|
Publication
|
|
Tuesday, 09 March 2010 17:28 |
|
T. Sucontphunt, B. Tunwattanapong, Z. Deng, and U. Neumann, Proceeding of Graphics Interface (GI) 2010, Ottawa, Ontario, Canada, May-June 2010, pp. 209-216. [pdf][bibtex][demo]
|
|
Last Updated on Friday, 30 July 2010 16:35 |
|
Portfolio
|
|
Thursday, 30 July 2009 03:59 |
|
These are some of the SubSurface Scattering Rendering results I implemented on PBRT framework for my final project of Advanced Computer Graphics: Photographic Image Synthesis taught by Prof. Paul Debevec, Prof. Pieter Peers, and Prof. Abhijeet Ghosh, Spring 2009. (This is a very good class). The whole milk dragon.  |
|
Last Updated on Wednesday, 09 September 2009 00:44 |
|
Read more...
|
|
Publication
|
|
Thursday, 21 February 2008 13:46 |
|
T. Sucontphunt, Z. Mo, U. Neumann, and Z. Deng, Proceeding of Graphics Interface (GI) 2008, Windsor, Ontario, Canada, May 28-30, 2008, pp. 177-184. [pdf][bibtex][demo ]  |
|
Last Updated on Tuesday, 09 March 2010 17:38 |
|
Portfolio
|
|
Tuesday, 11 December 2007 16:00 |
|
December 2007. This is a third person shooting game developed from XNA framework. I used pencil drawing style to render the game to make it a little different. Game play is very simple, trying to kill other opponent. The game starts being 2 players mode but player can change the mode to 1 player or 0 player (who would do that?) at anytime in the game, by set either player to be "Auto-Monster". Player can change the controlled monster at anytime and the weapon will be changed according to the monster selection, and we have four monsters/weapons to be selected here. For each round, the scene will be changed, and we have three scenes. |
|
Last Updated on Tuesday, 21 April 2009 15:12 |
|
Read more...
|
|
Portfolio
|
|
Friday, 21 April 2006 19:56 |
|
Animation Project: Physically-Based Wind Force effects on Tree Animation. May 2003. Basically, I developed 3D animation of the Tree that can be bended by applying simulated wind force. The work focused on realistic interaction between wind force and the auto-generated L-Systems (Lindenmayer) tree structure. The main work is on animating realistic interaction between the objects (tree's branches) and the virtual environment (wind force) by using Forward Kinematics: Time Integration - Standard Euler method. This also makes objects interacting in real time without key-framing effort from animator. Each branch of the tree placed into this virtual environment has its own properties and the global environment has a job to apply the law of physics involved to each of them. Please follow the links for details. |
|
Last Updated on Saturday, 05 September 2009 10:50 |
|
Read more...
|
|
|
Portfolio
|
|
Friday, 21 April 2006 18:51 |
|
A Content-Based Clustering MetaSearch Engine. November 2000. Advisor: Dr. Thanwadee (Thanitsukkarn) Sunetnanta. In short, the work focused on internet search engine using hybrid clustering techniques of nonhierarchical clustering and a single link Hierarchical Agglomerative Clustering (HAC) with the search result documents. Basically, we used nonhierarchical clustering to group relevant documents together and used single link HAC to merge relevant groups once more. Additionally, the document’s content keywords are extracted using NRC’s Extractor technique. This technique helps documents in a group relate to the others more precisely. We also completed an empirical study to evaluate our experiment by adapting IR techniques of Average Precision, Precision Recall, and Expected Search Length measurement as a proof of our concept. The result shows that the sample searchers will spend less effort in finding relevant documents or sift through the irrelevant documents. User interface screenshot: 
Cluster results screenshot: 
|
|
Last Updated on Saturday, 05 September 2009 10:51 |
|
Read more...
|
|
|
|
|