Publications

• Can Voice Quality help Mandarin Tone Recognition? by Dinoj Surendran and Gina-Anne Levow, Proceedings of ICASSP 2008.
  
  
• Dialog Act Tagging with Support Vector Machines and Hidden Markov Models, Dinoj Surendran and Gina-Anne Levow, Proceedings of Interspeech 2006.
  
  
• Interactive Glyph-Supplemented Scatterplots for Visualizing Algorithm Results. Dinoj Surendran. (Interactive Demo.) Companion Volume to the Proceedings of HLT-NAACL 2006. [Reviewer Comments] [Project Page]
  
  
• "Quantifying the functional load of phonemic oppositions, distinctive features, and suprasegmentals", Dinoj Surendran and Partha Niyogi, book chapter in Nedergaard Thomsen, Ole (ed. fc.) Current trends in the theory of linguistic change. In commemoration of Eugenio Coseriu (1921-2002). Amsterdam & Philadelphia: Benjamins.
  
  
• Tone Recognition in Mandarin using Focus. Dinoj Surendran, Gina-Anne Levow, Yi Xu (Phonetics Department, UCL). Proceedings of the 9th European Conference of Speech Communication and Technology (Interspeech/ICSLP 2005)
  
  
• Visualizing high dimensional datasets with Partiview. Dinoj Surendran and Stuart Levy (NCSA, UIUC). Poster Proceedings of the Tenth IEEE Conference on Information Visualization, Austin, TX, 2004.
  
  
• The functional load of tone in Mandarin is as high as that of vowels. Dinoj Surendran and Gina-Anne Levow. Proceedings of Speech Prosody 2004, Nara, Japan, pp. 99-102.

Non-First-Author Publications

• "Title Classification of Statement and Question Intonations in Mandarin", Fang Liu, Dinoj Surendran, and Yi Xu. Proceedings of Speech Prosody 2006, Dresden, Germany.
  
  
• Estimation of Stellar Atmospheric Parameters from SDSS ugriz Photometry. Y. Lee, T. C. Beers, C. Bailer-Jones, H. J. Newberg, M. Subbarao, D. Surendran. Proceedings of the 207th Meeting of the Astrononomical Association of America, January 2006.
  
  
• Articulatory complexity, ambient frequency and functional load as predictors of consonant development in children.. Stephanie Stokes (Dept of Psychology, University of Reading) and Dinoj Surendran. Journal of Speech and Hearing Research 48(3): 577-91, June 2005.

Technical Reports

All are from the University of Chicago Computer Science Department unless otherwise stated.

• Local Rhyme-based Acoustic Features for Mandarin Tone Recognition. Dinoj Surendran and Gina-Anne Levow (TR-2006-05)

• Additional Cues for Mandarin Tone Recognition, Dinoj Surendran and Gina-Anne Levow (TR-2006-04)

• Measuring the usefulness (functional load) of phonological contrasts. (PDF version). Dinoj Surendran and Partha Niyogi (TR-2003-12)

Non-research publications are listed on the Writing page.

Presentations

Here are presentations I've been involved in. I was the one doing the talking, unless otherwise mentioned.

• May 2005: "Automating Visualization", at the workshop "Visualization of Astrophysical Data: Bringing Together Science, Art, & Education", Kavli Institute for Cosmological Physics, Chicago, IL.
• January 2005: "3d Visualizations of Current Astrophysical Research for Public Explorations and for the Classroom". Randy Landsberg, Mark SubbaRao and Dinoj Surendran. (Presented by Randy and Mark.) Presented at the Gadgets and Gizmos workshop at the 205th Meeting of the American Astronomical Society, San Diego, CA.
• December 2004: "Partiview: A Tool to Visually Analyze the Perfomance of Dimensionality Reduction and Clustering Algorithms". Dinoj Surendran and Stuart Levy (NCSA, UIUC). Demo at NIPS 2004.
• October 2004: "Visualizing High-dimensional datasets with Partiview". Dinoj Surendran and Stuart Levy. Joint presentation of an interactive poster at this year's Information Visualization conference in Austin, TX.
• October 2004: "Low-cost Stereo Visualization of Astronomical and Astrophysical Data". Randy Landsberg, Dinoj Surendran, Mark SubbaRao. (Presented by Randy and Dinoj.) Demo at the 4th Annual Great Lakes Planetarium Association Conference, Detroit, MI.
• September 2004: "Making Pretty Pictures for Machine Learning". Dinoj Surendran. Presentation at the Student Session at the Machine Learning Summer School, Berder Island, France.
• July 2004: "Visualizing the Sloan Digital Sky Survey in 3d Stereo". Mark SubbaRao and Dinoj Surendran. Joint presentation at the Physics Education and Outreach Workshop at the Aspen Center for Physics, Aspen, CO.
• March 2004: Cosmus: Virtual 3D Cosmology in Public Science Museums. "VR for Public Consumption" workshop at Virtual Reality 2004. Stephanie Andrews (presenter), Dinoj Surendran, Randall H. Landsberg, Eric Jojola, Leo Kadanoff, Ronen Mir, Joi Podgorny, Daniela Rosner, Mark SubbaRao.
• March 2004: "The functional load of tone in Mandarin is as high as that of vowels". Dinoj Surendran and Gina Levow. Poster at Speech Prosody 2004. Nara, Japan.
• November 2003. "The Quantification of Functional Load". Masters Thesis Presentation.
• August 2003: "Quantifying the Importance of Functional Load." Dinoj Surendran and Partha Niyogi. Paper presented at the 16-th Conference for Historical Linguistics, Copenhagen, Denmark.
• March 2002. "Language does not have to evolve optimally" Dinoj Surendran and Partha Niyogi. Fourth International Conference on Language Evolution, Harvard University, Cambridge MA.

Other Documents

Here are some short papers/memos/notes I've written while investigating some problems in speech / language.

• Memo: detecting obstruents
• Phonological and Syntactic information some quick calculations for John.
• Object Detection: a summary of some computer vision papers of Amit and Geman
• Total error as a product of band-passed error A test (which passes!) of Jont Allen's interpretation of Fletcher's Articulation Index rule, using data from Miller and Nicely

Not really research in the thesis sense of the word, just term papers. Well, I had to put them *somewhere* !

• How physiological properties of neurons emerge from more elementary ones
• Introduction to smartcards Fall 2000 Computer Architecture survey, with Ivona Bezakova and Oleg Pashko
• Working notes for presenting the 1999 paper "Statistical Models for Text Segmentation" by Beeferman/Berger/Lafferty in Levow's Discourse & Dialog class.

Confusion matrices

Here are several confusion matrices from psycholinguistic experiments done by other people in the past.

Code

Some code I've written. Sorry for the documentation, or lack thereof. The Matlab code is better documented than the C++ code. I meant to clean this up before releasing it, but that never happened...

• Matlab utilities. My other Matlab code often uses this stuff. So if you use any of my other Matlab code, put this in a place whose path is accessible by Matlab.
• A linear SVM (Mangasarian and Musicant's Lagrangian SVM) in 500 lines of C. Not quite 11 lines of Matlab, but it doesn't use any libraries! Oh, and here's a postscript file with details.
• PHONMAT a MATLAB class I've found invaluable for analyzing confusion matrices, etc. There is some documentation.
• PHONVEC The 1-dimensional version of PHONMAT. Also invaluable for functional load analysis.
• LABELS - a MATLAB class requird by PHONMAT and PHONVEC. For some documentation, go to the bottom of that for PHONMAT.
• Functional load C++ and Matlab code (Poorly documented, and not in full generality!). With a syllable-frequency list for English (Switchboard) .
• Matlab code for the ill-fated edge maps project (even more badly documented).

Other stuff

In the process of trying to finding a doctoral topic, I began a psycholinguistics project. Finishing it is on my List Of Things To Procrastinate, and will remain there for some time.