TeachingWs09MATDCV

Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Tracking and Detection in Computer Vision - Winter Term 2009/10

Administrative Info

Lecture by Dr. Slobodan Ilic
Exercises by: Stefan Hinterstoisser, Stefan Holzer

Type: Lecture Module IN2210
Programs: Informatics (Bachelor, Master)
Biomedical Computing (Master)
Robotics, Cognition, Intelligence(Master)
SWS: 2+2
ECTS: 5 Credits
Course Language: English

Time, Location & Requirements

Monday 10:00 - 11:30 MI 03.13.010

Exercises: Wednesday 10:00 - 11:30 MI 03.13.008

Consulting hours:
Lectures (Slobodan Ilic): Tuesday, 10:00-11:00h MI 03.13.040

Exercises: (Stefan Hinterstoisser and Stefan Holzer):

Requirements:

  • The classes and exam are in English
  • For the intermediate exam, nothing but your pens and your calculators are allowed
  • For the final exam, nothing but your pens, your calculators and one DIN A4 page (handwritten, front and back side) with notes are allowed
  • The final exam contains 100 points, you need to have 50 points to pass it
  • Up to 40 bonus points can be earned from the homework and the intermediate exam

Site Content

Announcements

Repeated Exam: Thu, 22.04.2010, 10:00-12:00, MI 03.13.010. Please send us an e-mail if you would like to take the repeated exam by 15.04.2010!

For the repeated exam we allow using of one A4 paper filled by hand.

Results for the final exams with grades: GRADES.

Consulting hours:

Lectures (Slobodan Ilic): Tuesday, 10:00-11:00h MI 03.13.040

Exercises: (Stefan Hinterstoisser and Stefan Holzer): Thursday, 10:00-12:00 MI 03.13.042

Access to the computerroom

Every student needs to have access to the computer in the computer room (03.13.008) before the exercises start! If you do not have access yet please sent us an email (send it to hinterst at in.tum.de ) in the following format:

username,name,family_name,tum_email_address

Registration

Neither the class nor the exercises require any registration. However, you need to be registered to the final exam via TUMOnline!

Homeworks

Students have exactly one week to solve the homework (Solutions will be checked in the following exercise session)

Overview

Computer Vision, as a relatively young research area, has emerged as a key discipline in computer science. This is not only evident by a growing high competitive research community with the high impact factor in computer science, but also by the emerging of numerous Vision companies turning the research ideas into the numerous commercial applications. Besides well know studies of the 3D geometry and camera models, object tracking and detection in the images and videos become one of the principal research directions of the modern Computer Vision. The main objective of this course is to provide students with gradual introduction to the modern tracking, detection and recognition techniques developed in last 5 years. The course will provide in depth knowledge of image features, their detection and description, matching techniques, key-point recognition, basic and advanced tracking algorithms based on image features and image intensities, basics of structure from motion and bundle adjustment, synchronous localization and mapping algorithm, object detection and pose estimation using classification approaches and object recognition and image retrieval using bag of words approaches. In the end student will have thorough description of the most important tracking, detection and recognition techniques. They should be able to understand and implement those solutions and apply them in the reasonably complex problems. The concepts described in this course will be accompanied with the brief explanation of the necessary mathematical tools. The participants of this course will be given the necessary basis to follow vast and growing Computer Vision literature and use the acquired knowledge to solve new practical problems.

Who should attend this course?

Bachelor students at their last years and all Master students that are interested to learn about the modern and well established concepts and algorithms related to tracking, detection and recognition of Computer Vision or desire to use those techniques in their research.
Prerequisites: Most of the knowledge required should be part of the normal background in Computer Science and undergraduate/graduate Mathematics and Geometry.


Textbooks

For some basic concepts the textbook "Computer Vision: a modern approach" by David Forsyth and Jean Ponce, can be used. However, since the majority of the topics is state-of-the-art, students will be referred to the relevant publications provided before the course and can use the provided lecture notes.

Lecture Schedule and Content

Lectures start on 19.10.2009 at 10am at MI 03.13.010.

Date Topic Conducted by Material/Literature
Mon, Oct 26 Introduction Dr. S. Ilic Introduction (without Videos)
Mon, Nov 2 and Nov 9 Filtering and edge detection: Convolution; Gaussians; Image derivatives; Canny edge detector Dr. S. Ilic Filtering and edge detection
Mon, Nov 16 Local invariant feature detectors: Harris corner detector; Scale space Dr. S. Ilic Feature detectors I - lectures
Local invariant features: Survey: T.Tuytelaars and K. Mikolajczyk
Mon, Nov 23 Local invariant feature detectors: Harris Laplace/Affine;EBR(edge based regions); Blob detectors; Hessian; Hessian-Laplace/Affine; Region detectors: MSER(Maximal Stable Extremal Regions), IBR (Image Based Regions); Dr. S. Ilic Feature detectors II - lectures
Related papers:
1. Harris-Laplace
2. Harris-Affine
3. Edge Based regions (EBR) and Intensity based regions (IBR)
4. Maximally Stable Extremal Regions (MSER)
Mon, Nov 30 Feature descriptors: Difference of Gaussians and SIFT; Integral images and SURF, Histogram of Oriented Gradients Dr. S. Ilic Feature descriptors
Related papers:
1. SIFT
2. SURF
3. HOG
Mon, Dec 07 Keypoint recognition: Randomized trees; FERNS; Keypoint signatures Dr. S. Ilic Keypoint recognition
Related papers:
1. Randomized Trees
2. FERNS
3. Signatures
Mon, Dec 14 Advanced methods for patch detection and pose estimation: Panter, Leopard and Gepard; DTT (Distance Transform Templates) S. Hinterstoisser and S. Holzer Example midterm exam questions
Mon, Dec 21 Camera models and projections; Model based tracking; Pose estimation from 2D-3D coresspondencies (DLT, P-n-P and POSIT); Rotation parametrization Dr. S. Ilic Camera models and pose estimation
Related paper:
Monocular Model-Based 3D Tracking of Rigid Objects: A Survey
Mon, Jan 11 Non-linear optimization; Robust estimators: RANSAC Dr. S. Ilic Non-linear and robust optimization
Related paper:
Monocular Model-Based 3D Tracking of Rigid Objects: A Survey
Mon, Jan 18, Jan 25 Template tracking methods: Lucas-Kanade,Compositional Alg., Inverse Compositional, ESM, Learning a linear predictor, Active Shape and Active Apperiance Models Dr. S. Ilic Template tracking methods: Lucas-Kanade,Compositional Alg., Inverse Compositional, ESM, Learning a linear predictor, ASM and AAA
Related paper:
1. Lucas-Kanade 20-Years On - A Unifying Approach
2. Hyperplane approximation for template tracking
Mon, Feb 1 Simulateneous Localization and Mapping(SLAM) Dr. S. Ilic Simulateneous Localization and Mapping(SLAM)
Related paper:
1. MonoSlam-Real-Time Single Camera SLAM
Mon, Feb 8 Summary and questions Dr. S. Ilic Summary of the course

Exercise Schedule

Exercises start on 11.11.2009 (due to FVV) at 10am at MI 03.13.008. Here you can find a MATLAB tutorial.

Date Topic Conducted by Material/Literature
Wed, Nov 11 Filtering and edge detection Stefan Hinterstoisser, Stefan Holzer Exercise01, Lena
Wed, Nov 18 Filtering and edge detection Stefan Hinterstoisser, Stefan Holzer Exercise02
Wed, Nov 25 No exercises due to internal event - however, there will be homework announced here (due on the 2.12.), Harris, Harris-Laplace Stefan Hinterstoisser, Stefan Holzer Exercise03, ICCV Paper
Wed, Dec 2 no homework (however there are exercises) due to dies academicus Stefan Hinterstoisser, Stefan Holzer ---
Wed, Dec 9 Sift, Dlt, Ransac, Image Stiching Stefan Hinterstoisser, Stefan Holzer Exercise04, image1, image2
Wed, Dec 23 Ferns Stefan Hinterstoisser, Stefan Holzer Exercise05,image sequence
Wed, Jan 13 Non-linear optimization Stefan Hinterstoisser, Stefan Holzer Exercise06,image sequence
Wed, Jan 20 Template Tracking (due date extended to 3.2.10!!!) Stefan Hinterstoisser, Stefan Holzer Exercise07
Wed, Feb 3 Kalman Filter Tracking Stefan Hinterstoisser, Stefan Holzer Exercise08, Ellipse, computeAnalyticMeasurementMatrices.m, evaluateAnalyticMeasurementMatrices.m


TeachingForm
Title: Tracking and Detection in Computer Vision
Professor: Slobodan Ilic
Tutors:  
Type: Lecture
Information: 2+2 SWS, 5 ECTS, Wahlfach im Masterstudiengang Informatik, Wahlfach im Masterstudiengang Biomedical Computing, Wahlfach im Masterstudiengang Robotics, Cognition, Intelligence
Term: 2009WiSe
Abstract:  


Edit | Attach | Refresh | Diffs | More | Revision r1.72 - 12 Apr 2010 - 10:01 - SlobodanIlic

Lehrstuhl für Computer Aided Medical Procedures & Augmented Reality    rss.gif