OldEvents2008
From SpeechWiki
m |
|||
Line 81: | Line 81: | ||
== 12/10/07 2:30pm == | == 12/10/07 2:30pm == | ||
Final Exams | Final Exams | ||
+ | |||
+ | [[Category:Events]] |
Latest revision as of 04:25, 26 January 2009
|
8/27/07 2:30pm
Introductory meeting
9/3/07 2:30pm
Labor day
9/10/07 2:30pm
(Arthur) prinicpled stream weight tuning for factored observation models
9/17/07 2:30pm
Bryce Hant & Alwan. A psychoacoustic-masking model to predict the perceptionof speech-like stimuli in noise. Speech Communication 40 (2003) 291–313
9/25/07 1:00pm
(Harsh) Tracking in Video using a Time Series Transformation Learning approach
Hello everyone.
This presentation will be based on this CVPR 2005 paper. You really don't need to read the paper (unless you badly want to!). Besides, I will be using slightly different notation.
There's a sizeable amount of mathematics involved. Before reading this paper, I had to revise some linear algebra and learn about Reproducing Kernel Hilbert Spaces {RKHS} (functional analysis). I suppose most of you if not all of you, are already aware of these concepts.
However, if any of you wish to have a refresher of this material, here are some excellent resources (and very small documents too):
Functional Analysis > do go through this if terms like "complete space", "Banach space", "Hilbert space of functions" etc. seem new.
Linear Algebra > you can read this if you find some terminology in the Functional Analysis document seen-before-but-quite-a-while-ago.
Finally, I am hoping that I will be able to devote minimal time to explaining what an RKHS is. Here's a document that explains it beautifully...only the first few pages are actually needed for understanding RKHS (pages 14-17 mainly)
Wahba - An Intro to Model Building with RKHS
If this seems too much, you can alternatively go through my slides on RKHS (these are also on the actual presentation).
In any case, if you wish I can go over the RKHS slides during the presentation (at the moment though, I'm not planning to).
10/1/07 2:30pm
No meeting - Mark's out of town.
10/8/07 2:30pm
(Jui-Ting Huang) One-class classification problems
10/22/07 2:30pm
Bryce practices his preliminary exam proposal.
10/29/07 2:30pm
Brief Review: Latent Semantic Analysis: towards insight into human cognition
Xiaodan
Ref:
S. Deerwester, Susan Dumais, G. W. Furnas, T. K. Landauer, R. Harshman (1990). "Indexing by Latent Semantic Analysis". Journal of the Society for Information Science 41 (6): 391-407. http://lsi.research.telcordia.com/lsi/papers/JASIS90.pdf
Mark Steyvers(2007). Probabilistic Topic Models. In T. Landauer, D. McNamara, S. Dennis & W. Kintsch (Eds.), Handbook of latent semantic analysis (pp. 467-480). Mahwah, NJ: Erlbaum. http://psiexp.ss.uci.edu/research/papers/SteyversGriffithsLSABookFormatted.pdf
11/5/07 2:30pm
Topic : Likelihood-Maximizing Beamforming
"One page summary for this topic"
"Likelihood-Maximizing Beamforming for Robust Hands-Free Speech Recognition" by M. L. Seltzer, B. Raj, and R. M. Stern, IEEE Trans. on Speech and Audio Processing, 2004.
"Subband Likelihood-Maximizing Beamforming for Speech Recognition in Reverberant Environments" by M. L. Seltzer and R. M. Stern, IEEE Trans. on Audio, Speech, and Language Processing, 2006.
11/12/07 2:30pm
Arthur practices his preliminary exam proposal.
11/19/07 2:30pm
Thanksgiving Break
11/26/07 2:30pm
12/3/07 2:30pm
12/10/07 2:30pm
Final Exams