CLASS SEQUENCE (w adjustments in "real time")

- Intro and camera model

  • Motivation & Intro
  • Images & Point Clouds
  • Camera Model 
-  Geometric Backg (2D transformations)
  • LSE (homogeneous & Pseudoinverse) 
  • Camera calibration
  • Pose estimation (known K)
-Homography and 2D registration
  • DLT
  • Keypoints, 
  • Matching, Outlier Removal
- 3D Registration, 
  •  Procrustes Problem,
  • 3D registration ICP
- Stereo Vision
  •     Essential and Fundamental Matrix, 
  •     Triangulation, Reconstruction
  •     DLT and Robust estimation of F
-Structure From Motion
  • Tracking (the LK tracker)
  • Projective SfM pipeline and Tools (OpenSFM, Colmap)
  • Factorization based SfM
- Image processing
  • Linear/Nonlinear Filtering
  • Edge Detection
  •  Line Detection
- Recognition and Classification
  • Artificial neural networks
  • low level filters -> high level
  • feature maps
-Detection 
  • Region proposals
  • RoI pooling
  • Architectures
-Segmentation
  • Architecture structure
  • Semantic     segmentation andInstance segmentation
 - Special Topics
 - Seminar by a computer vision tech company (TBD)
Last class  quiz on project topics