http:^^www.cs.cornell.edu^info^projects^m-eng^
来自「This data set contains WWW-pages collect」· EDU^INFO^PROJECTS^M-ENG^ 代码 · 共 327 行
EDU^INFO^PROJECTS^M-ENG^
327 行
MIME-Version: 1.0
Server: CERN/3.0
Date: Wednesday, 20-Nov-96 19:52:59 GMT
Content-Type: text/html
Content-Length: 12147
Last-Modified: Monday, 02-Sep-96 00:56:57 GMT
<html><head><title>M-Eng Projects</title></head><body><h1><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><!WA0><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/Car-big.jpg"><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><!WA1><img src="http://www.cs.cornell.edu/Info/Projects/M-Eng/Car-sml.jpg" align=left vspace=10></a><br>Masters of<br>Engineering<br>Projects</h1> <br clear=left>This place is a first attempt at collecting M-Eng projects. The goal isto provide a show-case for sucessfully completed projects as well asideas for new projects. Contributions are welcome...<p><h1>This page is from 1995! Some of the projects might still be available.</h1><hr>A few reminders:<ul><li>This is an incomplete list. If you have a specific topic in mind, feelfree to contact the appropriate faculty member directly.<p><li>Many of the projects can be adapted from 3 to 6 credits depending onhow much work you can put in.<p><li>You need to sign up 'til sept. 30th if you want to get project creditsthis semester. <b>However</b>, there is generally no reason to. You cando the project now and get credit in the spring...<p><li>If you find that most of the interesting projects require knowledgeyou don't have, look into extending a course project into the M-Engproject. That's the best way to get the background.<p></ul><H2>Project proposals</H2><h3>Claire Cardie</h3><ul><li>Build any of a number of statistical models of languageprocessing to be used for tasks such as part-of-speech tagging andword sense tagging of natural language text. We will evaluate themodels by testing their performance on a corpus of real-world textand comparing the results to the performance of an existing languagelearning system that relies instead on symbolic machine learningtechniques. Note: No knowledge of linguistics is required.<p><li>We have a language acquisition system that uses symbolic machinelearning techniques to perform a number of disambiguation tasks onnatural language texts at the word level (e.g., decide theappropriate meaning or part of speech of a word in a given context).This project would involve extending the system to handle any of anumber of higher level disambiguation tasks like (1) deciding whatphrase a prepositional phrase modifies (e.g., "I went to the storeIN JUNE" vs. "I went to the store IN ITHACA"), or (2)distinguishing metaphorical from literal uses of a phrase (e.g.,deciding that "...KILLING two birds with one stone" does not imply areal murder). Note: No knowledge of linguistics is required. <p><li><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><!WA2><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_doc.html">Docu-Vault project</a>: the file cabinet ofthe future.</ul><h3>Nikos Chrisochoides</h3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><!WA3><AHREF="http://www.tc.cornell.edu/Research/ACRI/People/Nikos/Projects/index.html">A number of projects on Problem Solving Environments</A><h3>Robert Constable</h3>Any project I offer will be part of CS572 and willassume knowledge of Nuprl taught in that course.I will have several ideas for projects during thecourse.<h3>Caroline Hecht</h3>A number of <!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><!WA4><a href="http://www.tc.cornell.edu/MEng/">projectsare available in the Theory Center</a>.<h3>Thomas Henzinger</h3> For the following projects, you should be familiar with C and C++ programming, finite automata theory, a little bit of logic, and college level calculus. The additional requirements for each project are listed below. <ul> <li>Hybrid System Simulation: Write a simulator for hybrid systems, which are distributed real-time systems that consist of interactingdigital and analog components.<li>User Interface: Write a parser for a graphical hybrid system description language, which provides the input for the hybrid system simulator. You have to know X-window programming. <li>Scientific Visualization: Use computer graphics and animationtools to visualize the output of the hybrid system simulator.You have to know X-window programming, some computer graphics,and an animation tool. <li>Programming Language Translation: Write a compiler to translate one hybrid system description language into another. You have to know lex and yacc.</ul><h3>Bruce Land</h3>Possible projects:<ul><li>visuailzation of parallel algorithms<li>DSTool interface to DX<li>greyscale stereogram by cycling through several b/w stereograms<li>Mosaic document structure browser<li>Compress/decompress modules for network computing<li>Mosaic conversion of CS418/417 documents<li>Parallelize tensor modules-eigenvalue/vector, arithmetic<li>Matrix inversion module<li>High dimensional vis development -- rotations, figure id, correlations,hyperplanes<li>Neural simulator module, single compartment, nonlinear cell or abstractneuron<li>Robot vision/pattern recognition modules<li>Vector field texture calculation --SIGGRAPH93<li>Surface integral module -- sum over S of (V dot n)<li>Oriented particles -- SIGGRAPH 92, page 185<li>Sea shell modeling -- SIGGRAPH 92, page 379<li>Evolution for graphics -- SIGGRAPH 91, page 319<li>Extraction of geometry from volume data -- SIGGRAPH 91, page 217<li>Particle animation using data parallel methods -- SIGGRAPH 90, page 405<li>Surfaces and spherical harmonics -- IEEE Computer Graphics July 88<li>Distributed vol rendering -- IEEE computer graphics, nov 1993, p16<li>Distributed wave computation -- SC'93 proceedings, p22<li>Parallel rendering -- SC'93 proceedings, pp 690, 699, 709<li>Antialias filter module for image output (convolution, also SIG91 page133 \& SIG94 page 59)<li>NTSC filter module<li>Psych application visual stimuli<li>HDF to DX \& DX to HDF data translator, particularly image format<li>DX to Mathematica using Mathlink from Wolfram Research<li>DX to Matlab<li>DX module to perform surface construction from sample points<li>Wavelet transform</ul><h3>Joe Skovira</h3><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><!WA5><a href="http://www.tc.cornell.edu/~skovira/MEng.html>Parallel JobScheduling.</a><h3>Brian Smith</h3><h3>Chunguang Sun</h3><h4>Parallel Graph-Theoretic Algorithms for Sparse Matrix Factorizations</h4>We are interested in the design and implementationof scalable parallel algorithms for three fundamentalsparse matrix problems: sparse Cholesky factorizationsparse LU factorization and sparse QR factorization.A typical sparse matrix factorization algorithm consists of a symbolic phase and a numeric phase. In the symbolic phase,the adjacency graph of a sparse matrix is reordered and the symbolic structure of its matrix factors is determined.In the numeric phase, the numerical values of the matrix factors are computed. While several efficient parallel algorithms for the numeric phase have recently been proposed,efficient and practical parallel algorithms for the symbolic phase are yet to be developed. Among the projects in this area are:<ul><li>Develop parallel algorithms for reordering a sparse matrixand implement them on the IBM SP2.<li>Develop parallel algorithms for structural predictionin a sparse matrix factorization and implement them on the IBM SP2.</ul>If you are interested in these projects, pleasecontact Chunguang Sun, 729 Rhodes Hall(csun@cs.cornell.edu).<h3>Steve Vavasis</h3><a href="http://www.cs.cornell.edu/Info/People/vavasis/qmg-home.html">QMG</a>is a C++ software package for finite element mesh generation.I have several projects connected with QMG including graphics,interoperation with other software packages, and extensions tothe basic algorithms. Prerequisites for this project are knowledgeof C++, matlab, and a numerical analysis course (e.g. CS421).<h3>Thorsten von Eicken</h3><ul><li>A 155MBit ATM network with 8 Sparc-20's is available for<!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><!WA6><a href="http://www.cs.cornell.edu/Info/Projects/ATM/m-eng.html">projects</a>. Some of thesinvolve parallel programming for which the IBM SP-2 is also available.<p><li><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><!WA7><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_www.html">WWW project</a>: develop extensions to WWW servers,protocols, and browsers to enhance the support of video, interactivity, customizability, etc...<p><li><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><!WA8><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_doc.html">Docu-Vault project</a>: the file cabinet ofthe future.</ul><h3>Werner Vogels</h3><ul><li><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><!WA9><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_airsim.html">Air-traffic simulator project</a>: buildthe various blocks that make up a real air-traffic control system sothat we can construct an experimental environment for the nextgeneration air traffic control systems.<p><li><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><!WA10><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_cscw.html">Horus groupware project</a>: build acollection of small modules using Horus that will assist the softwaredesigner in building a wide variety of groupware applications.<p><li><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><!WA11><a href="http://www.cs.cornell.edu/Info/Projects/M-Eng/prop_os.html">(Distributed) Operating Systems research projects</a>To give students the opportunity to get their hands dirty onsome <EM>real</EM> operating systems research we are offeringa number of projects based on the<!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><!WA12><A HREF="http://plan9.att.com/plan9"> plan9 </A>distributed operating system and on the<!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><!WA13><A HREF="http://www.cen.uiuc.edu/~jeske/VSTa/"> VSTA </A>public micro kernel.</ul><h3>Ramin Zabih</h3><ul><li>Write a package for displaying images and color overlays, probably on top ofX11R6 or OpenGL.<li>Build a system for finding and tracking people in a sequence of images.Experience with C and images required.<li>Experiment with the use of visual motion for image compression.<li>Write C programs that write C programs for extremely efficient local imageprocessing. Experience with images not required, but extensive experiencewith C (and perhaps some knowledge of Lisp) is necessary.<li>Construct and benchmark algorithms for visual motion and stereo based onrobust statistics (especially: least median of squares). Minimal experiencewith images is required.<li>Various <!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><!WA14><ahref="http://www.cs.cornell.edu/Info/Projects/csrvl/csrvlMeng.html">projects</a>in the Robot-Vision Lab.</ul><h3>Rich Zippel</h3>There are two basic areas in which we are active, development of very highlevel software development tools, especially in scientific computation, andinfrastructure to support collaboration among people separated by time andspace.In the very high level software area we are developing a new programmingparadigm, where very high level specifications of computations areconverted to executable programs via a sequence of user controlled programtransformations. Among the projects in this area are:<ul><li>Develop and code new program transformations (e.g. forparallelization, new numerical computation techniques, or transformationthat instantiate a particular concurrency control mechanism).<li>By couple the existing transformations and external tools, performsome large scale simulations, e.g. bone growth, explosion of a comet,gravitation collapse of a pair of black holes.<li>Develop and improve the user interface to the program transformationprocess using either Apple Common Lisp, Unix Emacs or the SynthesizerGenerator. Since this our programs include descriptions of threedimensional objects this may include some novel interactions with graphicsprograms.<li>Linkage of program transformation technology with the theoremproving and computational logic techology being developed by Prof.Constable's group.</ul>For collaboration, we would like to integrate our tools and technologieswith the emerging standards on the Internet. A few of these projects arelisted below, but there are a wide number of different possibilities.<ul><li>The manuals for our tools have been written in Framemaker. Forhighest quality printing, we would perfer to LaTeX. Develop tools andprocesses so that World Wide Web documents (HTML) can be mantained inparallel with the Framemaker documents. Develop automatic tools andprocesses for going back and forth between Framemaker and Latex.<li>A WWW manual should be organized and linked in a different fashionthan a printed manual. Develop mechanisms and guidelines to aid inmaintaining printed and on-line manuals concurrently.<li>Develop databases and links between Lotus Notes and WWW for running asoftware development project. How should bug tracking be handled,releases, versions of manuals, user's comments, demonstrations, etc.?</ul><hr>For further information about this WWW page contact <!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><!WA15><AHREF="mailto:tve@cs.cornell.edu">Thorsten von Eicken</A></html>⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?