http:^^www.cs.cornell.edu^info^courses^spring-96^cs664^projects.html
来自「This data set contains WWW-pages collect」· HTML 代码 · 共 73 行
HTML
73 行
MIME-Version: 1.0
Server: CERN/3.0
Date: Monday, 16-Dec-96 23:20:57 GMT
Content-Type: text/html
Content-Length: 3313
Last-Modified: Saturday, 09-Mar-96 21:37:39 GMT
<TITLE>Suggested Projects for CS664 </TITLE> <HEAD><H1>Suggested Projects for CS664 </H1></HEAD> <li> Robotics<ol><li><em> Edge Clustering</em> Develop algorithm for clustering ofedges that represent a single object.<li><em>Performance Simulations</em> Investigate how high-performance visionand image processing applications perform on modern architectures, visionapplications tend to use large data structures (images) that can exhibit bothgood and bad spatial locality. Ideally, one would have a processor simulatorthat would be configureable: cache size, cache associativity, memory size,processor speed, instruction set, etc. The simulator should be able togenerate statistics such as cache hit/miss ratio. Hopefully the simulatorcould be extended to simulate a Shared Memory Multiprocessor (SMMP). Startingfrom scratch will be very hard, it may be better to adapt an existing programto our needs. Starting points: <ahref="http://www.cs.washington.edu/research/arch/">UW</a>.<li><em>Image Databases</em> There is currently a project here at Cornell todevelop a general purpose image database that will allow "intelligent"searching for images based on a large number of image features. We needpeople to implement different measures, for basics such as texture patternmatching to face recognition. Starting points: <ahref="http://www.virage.com/vir-res/demos/">the competitors</a>.<li><em>Camera Gaze Control</em> Create or adapt an existing object trackingalgorithm so that the output can be used to control the pan/tilt positions ofa desktop camera. The robotics lab is developing a relatively-cheap pan/tilthardware for a standard CCD camera. We would like the user to be able toclick on a portion of the image and have the camera pan/tilt to keep theobject centered in focus. <li><em>Find a Cool Use for Stereo Vision</em> Existing programs allow us togenerate good depth maps at near-real time speeds (5-15 FPS). We would likesomeone to create a "cute" application using the robots or some other platformthat would use the depth maps to perform some task, such as navigation orcollision avoidance (these are not actually the same, currently the easiestway for the robot to navigate is for it to go until it "crashes" intosomething, back up a little, then turn). <li><em>Stereo Camera Calibration</em> Develop a robust, reliable procedurefor calibrating a pair of stereo cameras, using a feedback loop between theoutput of the cameras as they view a test pattern and their next adjustedview.<li><em>Text Segmentation</em> Develop a system that can detect text in animage or series of images. This is not the same as OCR, where one tries totake an image of printed text and recover the text; rather, you shouldconcentrate on how to find text in an image. Typical examples of texts thatone might want to identify from a video signal include the bylines on networknews and scores and player stats from sports games.<li><em>Human Face Segmentation</em> Basically, find the objects in a scenethat look like human faces. How many people are there? How certain is thisanswer? Starting point: <ahref="http://www.cfar.umd.edu/ftp/TRs/CVL-Reports-1993/TR3176-Sirohey.ps.Z">Univ. Maryland</a>, also work at CMU by Kanade's group.</ol> ⌨️ 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?