Docsity
Docsity

Prepare for your exams
Prepare for your exams

Study with the several resources on Docsity


Earn points to download
Earn points to download

Earn points by helping other students or get them with a premium plan


Guidelines and tips
Guidelines and tips

Computer Vision Course Announcements and Topics - Prof. David Jacobs, Study notes of Computer Science

Announcements for a computer vision course, including exam and review session dates, hints for the final exam, and a list of topics covered in the course such as boundary detection, modeling and algorithms, stereo, optimization, learning, graphics, image editing, biomedical engineering, and big data sets. The document also mentions various researchers and their contributions to the field.

Typology: Study notes

Pre 2010

Uploaded on 02/13/2009

koofers-user-dys-1
koofers-user-dys-1 🇺🇸

10 documents

1 / 39

Toggle sidebar

Partial preview of the text

Download Computer Vision Course Announcements and Topics - Prof. David Jacobs and more Study notes Computer Science in PDF only on Docsity! Announcements • Final: Thursday, December 15, 8am, here. • Review Session, Wednesday, Dec 14, 1pm, AV Williams 4424. • Review sheet with practice problems on-line. Hints for Final • Focus on core techniques/ideas: convolution, gradients, statistical modeling, 3D geometry, motion matrices and optimization methods we studied, and their use in vision. • Of course, other topics from course may show up. Modeling + Algorithms • Build a simple model of the world (eg., flat, uniform intensity). • Find provably good algorithms. • Experiment on real world. • Update model. Problem: Too often models are simplistic or intractable. Stereo • Modeling: Geometry, Photometry. – Perspective camera, known pose allows reconstruction. – Epipolar Constraint. – Ordering constraint – Match similar intensities. – Disparities similar in nearby pixels. • Algorithms – Straightforward Reconstruction. – Shortest path, Graph Algorithms. Where is Computer Vision Going? • More Data, Faster Machines => • More Interaction with Other Fields. • Fundamental Problems Remain State of the art method Boykov et al., Fast Approximate Energy Minimization via Graph Cuts, International Conference on Computer Vision, September 1999. Ground truth (Seitz) (Comaniciu and Meer) Learning • Recognition using supervised learning. – Given examples of an object – Use classifiers: eg, SVMs, Winnow, Boosting. • Grouping using unsupervised learning. – Eg., E-M • Probabilistic Modeling – Eg., Graphical models, texture, …. Graphics • Common interest in modeling reflection, light, 3D shape. • Image-Based Rendering. input depth image novel view Szeliski Figure 1: 3D tracking software developed at Digital Domain was used on nearly every shot of the movie Titanic. (From Vision in Film and Special Effects Doug Roble, Digital Domain) Biomedical Engineering • Segmentation – Identify organs to measure them. – Find tumors. • Tracking – Is a heart beating properly? Is there dead tissue? • Registration/Matching. – Positions of Tumors in Surgery. Figures by kind permission of Eric Grimson; further information can be obtained from his web site http://www.ai.mit.edu/people/welg/welg.html. Figures by kind permission of Eric Grimson; further information can be obtained from his web site http://www.ai.mit.edu/people/welg/welg.html. Figures by kind permission of Eric Grimson; further information can be obtained from his web site http://www.ai.mit.edu/people/welg/welg.html. Big Data Sets • Image Data Bases – Kodak, commercial data bases w/ tens of thousands of images. – Internet, with millions? • Satellite imagery (Petabytes). – Monitoring effects of climate change. • Custom Data Sets Tracking and Activity Classification (ae J Ea | vada snt.ry Ue JO Werderp tosds YY (q) (paeAino SUPpOoT sv.oure> AUPUI JO SUNSISUOD a4 avis [Teg JJOd y (2) °9 Amo] iq) (re) ‘Surddew sinpxo} JOYE [pow GE oy) JO SMOTA JUDSIOyPIG “FT oand1y $ = , ; ou ‘WI PLIOsTe VOnsSNsUOSS. PAYS [up OL ndur au st ySrpat. aouanbas opis [EUISLIO at] JO SouTEAy OMT “CT Qand1y ‘SuLoyy apnsed papoous-ApAnoy “07 aand14 CEB! suo] "SMOLA Poyetedas ATapia IWS wos UOSIod ve JO UONIILNsSUOSaY “FZ QInSLy Smart Thumbnails
Docsity logo



Copyright © 2024 Ladybird Srl - Via Leonardo da Vinci 16, 10126, Torino, Italy - VAT 10816460017 - All rights reserved