TU Wien:Computer Vision VU (Sablatnig)/Prüfung 2018-12-07

Aus VoWi
Zur Navigation springen Zur Suche springen

Fragen wie gehabt, zusätzlich kam noch:

Machine Learning[Bearbeiten | Quelltext bearbeiten]

Erkläre folgende Begriffe:

  • Supervised vs. Unsupervised Learning
  • Generalization
  • Gradient descent

Deep Learning[Bearbeiten | Quelltext bearbeiten]

  • Difference between the "classical" machine learning pipeline in computer vision and deep learning methods.
  • Explain autoencoders and how they are applied in deep learning.

Interest Points[Bearbeiten | Quelltext bearbeiten]

Explain:

  • Why repeatable?
  • Why scale - invariant?
  • Why locally information rich?

Reconstruction[Bearbeiten | Quelltext bearbeiten]

  • What is structure from motion?
  • Given Fundamental matrix between two cameras and a point x in one image, what can you say about the corresponding point in the second image?