Computer Vision 1 (Winter Term 2024/2025)

Overview

• Course (2/2/0) consisting of:
  • Lectures in TOE/0317/H (Mommsenstr. 12) on Mondays, 11:10–12:40
  • Exercise groups
    • In APB-E065 on Tuesdays, 13:00–14:30
    • In APB-E065 on Tuesdays, 14:50–16:20
    • In APB-E067 on Thursdays, 11:10–12:40
    • In APB-E067 on Thursdays, 13:00–14:30
  • Self-study
  • Final Examination
• Lecturer: Bjoern Andres
• Teaching Assistant: Jannik Presberger
• Enrolment (OPAL). Additional rules for enrolment may apply, depending on the study programme.
• Forum

Contents

• Lectures
  • Introduction
  • Color spaces
  • Operators on digital images
    • Digital images
    • Point operators
    • Linear operators (esp. convolution)
    • Non-linear operators
    • Edge and corner detection
  • Classification of digital images
    • Logistic regression
    • Deep learning (basics)
    • Pixel classification
    • Excursus: Minimum st-cut problem and maximum st-flow problem
  • Decomposition of digital images
    • Multicut problem
    • Local search algorithms
  • Semantic segmentation of digital images
    • Node labeling multicut problem
    • Local search algorithms
  • Multiple object recognition in digital images
    • Node labeling multicut problem
  • Single object recognition in digital images
    • Partial quadratic assignment problem
  • Multiple object tracking in digital images
    • Constrained multicut problem
  • Single object tracking in digital images
    • Coupled partial quadratic assignment problems
    • Algorithms
  • Linear and integer optimization for computer vision
    • LP relaxations and lower bounds
    • Simplex algorithm
    • Branch-and-bound
    • Branch-and-cut