CSCI 363 Vision--Spring 2023

    Exam 2 Review

    Home | | Schedule | | Assignments | | Lecture Notes

    Topics for Exam 2:
    This sheet is intended to help you prepare for the second midterm in this course. The exam will cover all readings and lectures up through Tuesday, March 28, and all homework through HW 5. While you are expected to know the material presented prior to the first exam, this exam will focus on the material covered subsequent to the first exam. The exam will be closed book and closed note, but you will be allowed to bring one 8.5 x 11 inch sheet of paper with your own notes on it (one side only) to use as a reference during the exam. The following list of topics should help you study for the exam.

    1. Spatial Frequency

      Spatial frequency components of an image
        Fourier Series
        Fourier Transform
        2D frequency spectra of sinewaves, plaids and checkerboards
      Psychophysics
        Contrast Sensitivity function
        Frequency Adapatation
        Spatial Frequency channels
        Frequency spectrum of a Gaussian
        Band-pass filtering--relationship to smoothing at different scales
      V1 cells and Spatial Frequency
        Gabor filters and their Fourier transforms
        V1 cells as Gabor filters
        Spatial Frequency columns
        V1 simple vs. complex cells--AC/DC responses
        V1 cells response to temporal frequency
        Relationship of speed, temporal frequency and spatial frequency
        Non-linearities of neurons

    2. Binocular stereo

      Geometric Basis of stereo vision
      Crossed and uncrossed disparity
      The correspondence problem
      Random dot stereograms (what they are and why they are a problem for stereo algorithms)
      Constraints for solving the correspondence problem:
        Smoothness
        Uniqueness
        Similarity
        Epipolar
      The Marr-Poggio algorithm
      The Marr-Poggio-Grimson algorithm
      Biology of Stereo vision
        Constraints used by human visual system
        Use of zero crossings by human visual system
        Types of neurons tuned to disparity:
          Near
          Far
          Tuned Excitatory
          Tuned Inhibitory

    3. Motion Detection

      Models of motion detection: Correlation, Reichardt, Gradient
      The Gradient Constraint equation
      The aperture problem
      Intersection of constraints technique for computing motion
      Measuring velocity along a contour
      The smoothness constraint
      Motion Energy Filters
        Motion as orientation in space-time
        Spatio-temporal filters
        Gabor filter in space-time
        Motion Energy filters from Quadrature pairs
        Motion Opponency
        Velocity extraction
        Motion Energy for 2D filters
        Velocity as a plane in spatio-temporal frequency space
      Motion Illusions
        Rotating spirals
        Barberpole illusion
        Reversed Phi illusion
        Fluted square wave (square wave minus fundamental frequency component)
      Computing perpendicular velocities using the Gradient method.
      Combining multiple measurements using error minimization
      Biological Motion Processing
        Dorsal and Ventral streams in extra-striate cortex
        Motion processing in V1
        Motion processling in MT
        Responses to gratings and plaids (component vs. pattern motion)
        Speed and direction selectivity
        Motion opponency
        Inhibitory surround in MT cells

    Home | | Schedule | | Assignments | | Lecture Notes

    Constance Royden--croyden@holycross.edu
    Computer Science 363--Computational Vision
    Last Modified: March 29, 2023
    Page Expires: March 29, 2024