Computer Vision 1 (Winter Term 2023/2024)

Overview

• Course (2/2/0) consisting of:
  • Lectures in ZEU/LICH/H on Mondays, 11:10–12:40
  • Exercise groups
    • In APB-E069 on Tuesdays, 13:00–14:30
    • In APB-E069 on Tuesdays, 14:50–16:20
    • In APB-E069 on Thursdays, 9:20–10:50
    • In APB-E069 on Thursdays, 11:10–12:40
    • online on Mondays, 14:50–16:20
  • Self-study
  • Final Examination
• Lecturer: Bjoern Andres
• Teaching Assistant: Holger Heidrich, Jannik Presberger
• Enrolment (OPAL). Additional rules for enrolment may apply, depending on the study programme.

Contents

• Lectures
  • Introduction
  • Operators on digital images (slides)
    • Point operators
    • Linear operators
    • Non-linear operators
    • Edge and corner detection
  • Classification of digital images
    • Logistic regression (slides)
    • Smooth pixel classification (slides)
    • Excursus: Maximum st-flow and minimum st-cut (slides)
    • Convolutional networks (slides)
  • Decomposition of digital images (slides)
    • Multicut problem
    • Algorithms
  • Semantic segmentation of digital images (slides)
    • Node labeling multicut problem
    • Algorithms
  • Multiple object recognition in digital images (slides)
    • Node labeling multicut problem
    • Algorithms
  • Multiple object tracking in digital images
    • Lifted multicut problem
    • Node labeling lifted multicut problem
    • Algorithms
  • Single object recognition in digital images
    • Quadratic assignment problem
    • Algorithms
  • Single object tracking in digital images
    • Coupled quadratic assignment problems
    • Algorithms