My solutions to exercises of Computer Vision course @ ETH Zurich

DLT, Golden standard, Camera model, Calibration

Harris corner, Feature descriptor, Feature matching

Particle filter, Bayes law, Localization

RANSAC line fitting, Fundamental matrix, Epipolar geometry, 8-point method, Adaptive RANSAC

Meanshift, Expectation Maximization (EM)

MVS, Disparity map, Winner-take-all local method, Graph-cut global method, Dense reconstruction

SfM, Initialization, Essential Matrix, PnP problem, Triangulation, Dense reconstruction

Template shape matching, Shape context, TPS transformation estimation

Shape from silhouettes

Histogram of Oriented Gradient(HOG), KMeans, Bag of Word, Nearest neighbor classification, Bayesian classification

CONDENSATION algorithm, particle filter, tracking

• Basic geometry and algebra
• Camera model
• Structure from motion
• Multiview stereo

• Harris corner
• SIFT
• Feature matching strategy

• Kalman filter
• Particle filter
• Optical flow

• EM
• Meanshift
• Graph cut based
• Deep learning based

• Template matching
• Sliding window and proposal
• Deep learning based

• Classifiers
• Bag of word
• Deep learning based