Computer Vision

I am interested understanding the things with mathematics. The world can be viewed as generating a set of signals, whether they be an image of someone you recognize like your brother or sister, the sounds generated by a human that you come to recognize as your name, the structure of a given protein string in our body, all of these are signals which contain patterns. These patterns can be described if we can effectively learn how to model the hidden variables that describe them.

Currently I am mainly working on trying to model the signals generated by the reflection of light on objects. We humans are particularly good at implicitly recognizing signals of this nature. How is it that we can recognize someone we know in an instant? How is it possible from me to recognize a kangaroo if I saw one, despite the fact that I have only seen of few pictures of these species in a book? In trying to find answers to these questions I have joined the Computer Vision lab at UCF.

 

• Automated Visual Inspection of Railroads

• Automated Surveillance System

• Object Detection and tracking

• Human Detection

• Implementing Lukas and Kanade’s Optical Flow
• Edge Detection
• A computational framework for the measurement of visual motion
• Implementation of Gaussian Pyramids
• Human Detection & Classification
• Image Segmentation using EM
• Object Matching in Disjoint Cameras using a Color Transfer Approach
• CVPR 2005
• Video Lectures (Thanks to the Mathematical Sciences Research Institute):
  • David Mumford, Pattern Theory: Grenander's Ideas and Examples
  • Andrew Zisserman, Stochastic Models & Learning I
  • Oliver Faugeras, Variational Principles and PDE's of Computer Vision
    
  • Trevor Hastie, Modern Classifier Design
    
  • Pietro Perona, An Invitation to Visual Recognition
    
  • Donald German, Strategies for Visual Recognition
    
  • Edward Adelson, Image Statistics and Surface Perception
    
  • Richard Baraniuk, Multiscale Geometric Analysis for Images
    
  • David Mumford, Modeling Shape
    
  • Oliver Faugeras, Variational Methods for Multimodal Image Matching: Theory and Applications
    
  • David Donoho, Appearance Manifolds 2
    
  • Luminita Vese, Energy Minimization for Cartoon & Texture Separation :U+V Models
    
  • Eero Simoncelli, Statistical Image Models
    
  • Bruno Olshausen, What We Know and Don't Know About Biological Vision
    
  • Song-Chun Zhu, Seeing as Statistical Inference
    
  • Jitendra Malik, Ecological Statistics of Grouping and Figure-Ground Cues
    
  • Oliver Faugeras, Variations on Image and Shape Warping, Statistics and Segmentation
    
  • Joachim Buhmann, Learning and Image Segmentation
• Source Code
  • Matlab
• Tutorials:
  • Kalman filters (De Schutter et al)
• Papers: Automatic Door Detection in Video Sequences
• Copyright:
  • Human Detection (United States Copyright Office, September 11, 2006)
    

 

Return to my home page

Please send comments to mikel at cs dot ucf dot edu