Class Note for EECS 841 with Professor Potetz at KU 9
Class Note for EECS 841 with Professor Potetz at KU 9
Popular in Course
Popular in Department
This 6 page Class Notes was uploaded by an elite notetaker on Friday February 6, 2015. The Class Notes belongs to a course at Kansas taught by a professor in Fall. Since its upload, it has received 18 views.
Reviews for Class Note for EECS 841 with Professor Potetz at KU 9
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 02/06/15
EECS 841 Computer Vision Brian Potetz Fall 2008 Lecture 23 Project Ideas EECS 841 Final Project What am I really looking for Find something that interests you Play around with it See if you can make it better Write about it Present it in class EECS 841 Final Project What am I really looking for Find something that interests you Play around with it See if you can make it better Write about it Present it in class Could be an application an algorithm an idea Lots of potential topic ideas to follow EECS 841 Final Project What am I really looking for Find something that interests you Play around with it See if you can make it better Write about it Present it in class Implement it make it work EECS 841 Final Project What am I really looking for Find something that interests you Play around with it See if you can make it better Write about it Present it in class How can you get the best performance How can you taylor an approach to a particular application Get creative But don39t get risky Start simple amp work upwards EECS 841 Final Project Grading 35 ofyour nal grade Writeup 75 Oral Presentation 75 Project content 20 Understanding ofthe material Quaity of results Creativity Thorough literature search Sticking to the proposal is encouraged See me if you get stuck EECS 841 Final Project Timeline Project Proposals Due Mon 102708 Written Projects Due Mon 120108 Oral Presentations 123 1210 No milestone reports are required But if you get stuck come talk to me emailappointmehtof ce hours Unexpected hangups are viewed more sympathetically if you come to me early EECS 841 Final Project Project Proposals Due Mon 102708 No need to wait until Monday I will provide feedback anytime Informal Couple of paragraphs You re not graded on it This is to help you Have some concrete ideas Know one or two papers in the area Some prior research will help you decide what issues are hard amp what problems are doable EECS 841 Final Project In return I will Let you know if the topic is appropriate Let you know if the scope is on target Provide you with some references to get you started Point out possible pitfalls EECS 841 Final Project The written report Around 6 pages for reasonable figure density Pretend it is a CVPR submission only shorter amp less ambitious httgcvgr2009orga uthorkitcvgrkitIgZ Don t sweat formatting details just follow the basic style Now is a good time to learn latex amp bibtex You are responsible for a thorough literature search EECS 841 Final Project The presentation 10 minute talks with 23 minutes for questions Provide a brief overview of the field But don tjust regurgitate the literature Talk about your own work Show your results Discuss limitationsproblems with the approach Don t cover every mathematical derivation Keep it high level High level topics from the projects will be on the nal exam I EECS 841 Final Project Other miscellaneous points Any language Matlab C etc is fine Implement your own code Use code from online only with my permission Matlab amp Image toolbox are OK Image Segmentation r39w 1V a I Llam Normalized Cuts Image Colorization httpwww cshuji acilyweissColorization Image Inpainting Parametric methods Fields of Experts Roth amp Black quotFields of Experts A Framework for Learning Image Priors 2005 15 Image Inpainting Nonparametric examplebased methods Cn39minisi Perez Toyama Region filling and object removal by exemplarbased inpainting 2004 16 Texture Synthesis uuuuuuuuuuwmurum ullsuabuwatuuy gig Atnda mars mum Tring momsquot as Hefthe fast nd itl was zonm39u as House Dquot as at nozaxs wnseas xiheli it lastnthzstbedian 941 Esamedqu n39 H m econicalHomdithAl Heftazs of as 62 Lewindail iian A1115quot as Leaving questits1astaticastica11He is dian Allastfal counda Lew atquotthis dai1yeas liin iedianicall Hoorewing roomsquot as House De ale f De Und itical coumestscxibed itIast fall He fall Hefft 39 39 1m 2 vinoI39m questica LewitL m icazs coecomsquot astoxeyeazs of Monica Lewinnw see a Thas Prinz loom smoniscatmwea u left a mouse bouestofMIe Idfta Le39st fastngine liuugsticazs Hef 1d it zipquot TrIouself a zingindits39onestldita ring quite astical cm s oxeyeaxs of Moung fall He ribof Mouse 4th left ringing questiol mavens of Monica Lewit ind Trippquot Thatnowsezr Political comedian l anr me years ofanda Trippquot That hedian A1 Lest fasee yea nia Tripp39iolitica1 comedianale39the fzwse ring qua nlitical com years of the smreaxs ofa39s l Frat nica L ms Lew se lest rim 1 He fas questnging of at beou Nonparametric examplebased methods htt ra hicscscmuedu e0 le efros research EfrosLeun htm Single Image SuperResolution Freeman Pasztor Carmichael quotLearning LowLevel Vision 2000 18 Single Image SuperResolution Freeman Pasztor Carmichael quotLearning Low Leve Vision 2000 Image Denoising Image Denoising Image Denoising PortiIIa et a Spectral Matting c Alpha matte b Hard segmentation a Input image Levin Rav Acha Lischinski quotSpectral Matting CVPR 2006 Spectral Matting Painted Poorly scanned htt raicswashin tonedu ro39ects uer Adaboost Car detector UIUC Image Database for Car Detection This database consists of 550 car training images 500 noncar training images approximately 17 0 test images in which cars are present at the same scale as in training and approximately 17 0 test images in which cars are present at more than one scale lquot Space Carving Kutulakos amp Seitz A Theory of Shape by Space Carving 2000 28 i5iuot oziyr sih htt hototourcswashin tonedu httpZiveabscomphotosynth Photo Tourism amp PhotoSynth htt iveabscom hotos nth htt hototourcswashin tonedu Nevertheless some real progress has been made G a mes PlayStation 2 EyeToy Sega SuperStars Vinua Fighter Wexler Shechtman amp Irani quotSpaceTime Video Completionquot CVPR 2004 33 Spacetime video completion In InnIx sequence 1 2 s 1 2 ll llmludh lemma Wexler Shechtman amp Irani quotSpaceTime Video Completionquot CVPR 2004 Zelnik Manor amp Irani httpwwwwisdomweizmannac iImathusersvisionVideoAnalysisDemos EventDetectionEventDetection html 34 Other Ideas for Projects Ideas related to your own research It has to be computer vision Books Forsyth amp Ponce quotComputer Visionquot Shapiro amp Stockman quotComputer Visionquot Davies quotMachine Vision Theory Algorithms Practices Here is a list of project ideas from another course httpwww and rew cmueducourse167202008projectshtml 35 Have Fun
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'