Topics on Vision in Humans and Machines
CSE 291-B100, Spring 2005
David Kriegman
Kriegman {at} cs.ucsd.edu
Marian Stewart Bartlett
Marni {at} salk.edu
Grading: S/U, 2 units only.
This graduate seminar will explore topics in vision such as stereo, segmentation, motion interpretation, object recognition, face identity and expression recognition. In each class session, a topic will be considered from the perspectives of computer vision and human visual perception. There will be assigned readings on both human and machine vision, and a pair of students will serve as discussants from the two perspectives.
Class structure: Every week, each student will submit two questions or comments on the readings to the Discussion Board. Students will also complete tutorials from the Cold Spring Harbor course on Computational Visual Neuroscience.
3/30 Organizational Meeting, Center Hall 223 9:30-11.
4/06 Statistics of Natural Images
4/13 Information Theory and Neural Representation
4/20 Luminance, Illumination, and Contrast Normalization
4/27 Motion
5/4 Color
5/11 Texture/Segmentation
.
5/18 Object Representation and Recognition
5/25 Face Processing
6/1 Active Exploration
Reading List
4/06 Statistics
of Natural Images
Field (1987) Relations between the statistics of natural images and the response properties of cortical cells. J. Opt. Soc. Am. (or more recent papers, e.g. ‘What is the goal of sensory coding’ Neural Computation). pdf
J. Huang and D. Mumford. Statistics of natural images and models. CVPR, pages 541–7, 1999. pdf
Optional:
Ron O. Dror, Thomas K. Leung, Edward H. Adelson, and Alan S. Willsky,
Statistics of Real-World Illumination CVPR 2001 pdf
Ruderman & Bialek (1994). Statistics of natural images: scaling in the woods. Physical Review letters. pdf
Field (1999). Wavelets, vision and the statistics of natural
scenes. Phil. Trans. R. Soc. London A. 357, 2527-2542. pdf
4/13 Information
Theory and Neural Representation
Simoncelli & Olshaussen (2001). Natural image statistics and Neural Representation. Ann Rev Neurosci. This was on the reading list for the cogsci200 on learning in vision. pdf
Linsker (1990). Perceptual neural organization – Some approaches based on network models and information theory. Annual Review of Neurology 13, p. 257-281. pdf
Optional:
Barlow (1989). Unsupervised learning. (Or his original 1961 paper ‘possible principles underlying the transformations of sensory messages.’ It comes highly recommended.) pdf
Olshaussen & Field (1996). Emergence of simple cell receptive field properties by learning a sparse code for natural images. Nature 381, p. 607-609. pdf See also Vision Research 1997. pdf
Bell & Sejnowski (1996). The independent components of natural scenes are edge filters. Vision Research. pdf
Edelman & Intrator (2004).
Unsupervised statistical learning in vision: Computational principles,
biological evidence. Interesting but short. pdf
Shannon, C.E. (1948). A
Mathematical Theory of Communication. Bell System Technocal Journal 27,
379-423, 623-656. pdf
4/20 Luminance,
illumination, and contrast normalization
Wainwright, Schwartz & Simoncelli (2002). Natural image statistics and divisive normalization: Modeling nonlinearity and adaptation in cortical neurons. In Rao, Olshausen & Lewicki (eds). Probabilistic models of the brain: Perception and Neural Function. pdf
J. Ho, D. Kriegman, ``On the Efect of Illumination and Face Recognition” to appear in “Face Processing: Advanced Modeling and Methods,” R. Chellappa, W. Zhao, editors, 2005. pdf
Optional:
Adelson (2000). Lightness perception and lightness
illusions. In Gazzaniga (ed). The new cognitive neurosciences. pdf
Demos of Adelson’s lightness illusions.
4/27 Motion
Fleet, D. J., Black, M. J. and Nestares, O., (2002). Bayesian Inference of Visual Motion Boundaries, in Lakemeyer, G. and Nebel, B. (Eds.) Exploring Artificial Intelligence inthe New Millennium, Morgan Kaufmann Pub. pdf
J. Xiao, S. Baker, I. Matthews, and T. Kanade (2004). Real-Time Combined 2D+3D
Active Appearance Models. CVPR. pdf
Demos of the system in the Xiao paper and also demos of other AAM systems by the same group
Optional:
Albright (1993)
Cortical Processing of Visual Motion. In F.A. Miles and J. Wallman, Eds. Visual
Motion and its Role in the Stabilization of Gaze. Elsavier Science Publishers,
pg. 177-201. pdf
Adelson & Bergen (1985). Spatio-temporal energy models for the perception of motion. JOSA. pdf
Weiss, Simoncelli & Adelson (2002) Motion Illusions as optimal percepts. Nature Neuroscience. pdf
Black, M. J. and Anandan, P., (1996) Computer Vision and Image Understanding, CVIU, 63(1), pp. 75-104. pdf
Bruhn, Weikert and Schnorr (2005). Lucas/Kanade Meets Horn/Schunk: Combining Local and Global Optic Flow Methods. IJCV 61(3):211-231, Feb 2005. pdf
5/4 Color
von der Twer, T. and Macleod. D.I.A. (2001). Optimal
nonlinear codes for the perception of natural colors. Network, Computaiton
inNeural Systems, 12 395-407. pdf
NOTE: We
originally listed the following book chapter, but the Network paper is a little
easier to follow, so Josh recommends reading that one instead. If you read the
book chapter, then Josh will be
focusing on pgs 1-21 of the chapter, which covers the same material as the
Network paper. MacLeod, D.I.A. (2003). Colour Discrimination, Colour
Constancy, and Natural Scene
Statistics (The Verriest Lecture). In: J.D. Mollon, J. Pokorny, & K. Knoblauch (Eds.),
Proceedings of the Thomas Young
Symposium of the International Colour Vision Society London: Oxford
University Press. pdf
Neitz, Carroll, Yamauchi, Neitz, & Williams (2002).
Color Perception Is Mediated by a Plastic Neural Mechanism that Is Adjustable
in Adults. Neuron, Vol. 35, 783. pdf
Optional:
Brainard Kraft & Longere (2003). Color contstancy:
developing empirical tests of computational models. In Maunsfeld & Heyer
(Eds.) Colour Perception: Mind and the physical world. pdf
Delahunt & Brainard (2004). Does human color constancy incorporate the statistical regularity of natural daylight? Journal of vision 4 p. 57-81. pdf
Rhea Eskew. Chromatic Detection and Discrimination, In Gegenfurtner and Sharpe (Eds), Color Vision from Genes to Perception. Cambridge University Press. p. 345-368.
Nathans, J., (1989) The genes for color vision. Sci Am,
1989. 260(2): p.
42-9. pdf
5/11 Texture
/ segmentation
Malik & Perona (1990) Preattentive texture discrimination with early vision mechanisms. JOSA pdf
Zhu, Wu, & Mumford (1997). Minimax Entropy Principle and its Application to Texture Modeling. Neural Computation 9(8), p. 1627-1660. pdf
Optional:
Rosenholtz (2001). Search asymmetries? What search asymmetries? Perception & psychophysics. pdf
Rosenholtz (2000). Significantly different textures: a computational model of pre-attentive texture segmentation. ECCV. pdf
Farid & Adelson (2001). Synchrony does not promote grouping in temporally structured displays. Nature Neuroscience 4(9) p. 875-876. pdf
Heeger and Bergen (1995). Pyramid-Based Texture Analysis and Synthesis. Siggraph. P. 229-238. pdf
Fowlkes & Malik. Learning edge detection from human examples.
5/18 Object
representation
Kersten, Mamassian & Yuille (2004). Object perception as
Bayesian inference. Annual review of Psychology. pdf (See also Kersetn & Yuille 2003, Bayesian
models of object perception. Current Opinion in Neurobiology. ) pdf
Rothganger, F., Schmidt, C., Ponce, J. (2005). 3D Object Modeling Using Local Affine-Invariant Image Descriptors and Multi-View Spatial Constraints. International Journal of Computer Vision. pdf
Optional:
Edelman & Duvdevani-barr (1997). A model of visual recognition and categorization. Phil Trans. Royal Soc. London 352 p. 1191-1202. pdf
2D vs 3D (Mike Tarr) and/or Biederman & Kolocsoi
5/25 Face
Representation
Bartlett (submitted). Information maximization in face processing. Neurocomputing. pdf
Webster, Werner, & Field (in press) . “Adaptation and the phenomenology of perception.” In Clifford and Rhodes (Eds.) Fitting the mind to the world: adaptation and aftereffects in high level vision. (Mike Webster has some cool work on face adaptation that speaks to the information maximization idea.) doc
Optional:
Valentine (2000). Face Space models of face recognition. In Wenger & Townhend (Eds.) Computational, geometric, and process perspectives on facial cognition: Contexts and challenges. pdf
6/1 Active
Exploration
Sprague, Ballard and Robinson (submitted). Modeling
attention with Embodied Visual Behaviors.
ACM Transactions on Applied Perception. pdf
Walker-Renninger, Coughlan, Verghese, & Malik (2005). An
Information
Maximization Model of Eye Movements. NIPS 17, 1121-1128. pdf
Please also read the introduction of Nelson (in press) Psych
Review, for a review of probabilistic accounts of information acquisition (pgs
1-11 and 42-44, and the tables on 52,53,64). Nelson, J.D. (in press) Finding
useful questions. On Bayesian diagnosticity, probability, impact, and
information gain. Psychological Review. pdf
Optional:
Walther, D. , Rutishauser, U.,
Koch, C., and Perona, P. (in
press). Selective visual attention enables learning and recognition of
multiple objects in cluttered scenes. Computer Vision and Image Understanding. pdf
Denzler, J. & Brown, C.M.
(2002). Information theoretic sensor data selection for active object
recognition and state estimation. PAMI 24 p. 145-157. pdf
Yuille, Coughlan & Kersten
(1998). Computational vision: Principles of perceptual inference. Book
Chapter. This is a good chapter
introducing Bayesian models, as well as how they tie into information theory.
Also covers active exploration. pdf
Lee & Yu (1999). An Information-Theoretic Framework for
Understanding Saccadic Behaviors. NIPS 12, 834-840. pdf
Najemnik, J. & Geisler, W. S. (2005, March 17). Optimal eye movement
strategies in visual search. Nature, 434, 387-391. pdf
Nelson & Cottrell (submitted). A probabilistic model of
eye movements
in concept formation.
Related Tutorials
Cold Spring Harbor Computational Vision Tutorials by Eero Simoncelli, Paul Glimcher, EJ Cichilnisky, and others.
NIPS Computational
Vision Tutorial by Dan Kertsen
ICA Tutorial by Hyvarinen & Oja
Linear
Algebra tutorial by Eero Simoncelli
Linear Algebra in Parallel Distributed Processing by Mike Jordan.