Skip to content

CVML Exam Prep

Study notes, worked solutions, and cheat sheets for Computer Vision and Machine Learning (CVML) — TU Braunschweig, SoSe 2025.

All mathematics on this site is typeset with MathJax, so formulas such as the depth from disparity \( Z = \dfrac{b\,f}{d} \) or the Harris response \( R = \det(H) - k\,\operatorname{trace}(H)^2 \) render as real equations rather than plain text.

Chapters

# Topic Focus
01 Image Acquisition Pixels, colour spaces, pinhole, depth of field, distortion, noise
02 Filtering, Edges & Thresholding Convolution, Gaussian/Sobel/LoG/DoG, Otsu, adaptive thresholding
04 Machine Learning Perceptron, CNN, gradient descent, confusion matrix
05 Features Pinhole projection, Harris, SIFT, ZMNCC
06 Optical Flow Optical-flow equation, Lucas–Kanade, Horn–Schunck, FlowNet
07 Parametric Transformations Translation/rotation/scale, homography, splines, TPS
08 Epipolar Geometry & Depth \( Z = b\,f/d \), fundamental matrix, rectification, block matching
09 Morphing & View Synthesis Cross-dissolve, morphing, view-morph formula
10 Camera Calibration \( K \) matrix, ( [R\,
11 Neural Radiance Fields NeRF MLP, ray marching, positional encoding

(There is no Chapter 03 — the lecture set merges "Filtering" and "Edges & Thresholding" into Chapter 02.)

Exam resources

Reference

About the notes

Each chapter follows a 17-section rubric: Overview → Basics → Mathematical Foundations → Section notes → Code examples → Exercises → Trial-exam mapping → Algorithms → Applications → Summary → Hard questions → Full answers → Coding tasks → Mini project → Cheat sheet. Explanations are bilingual (English + বাংলা).