Download:
by W. T. Freeman, W. T. Freeman, J. B. Tenenbaum, J. B. Tenenbaum
http://www.merl.com/papers/docs/TR96-37.pdf
Add To MetaCart
Abstract:
In many vision problems, wewant to infer two #or more # hidden factors which interact to produce our observations. We may want to disentangle illuminant and object colors in color constancy; rendering conditions from surface shape in shape-from-shading; face identity and head pose in face recognition; or font and letter class in character recognition. We refer to these two factors generically as #style " and #content". Bilinear models o#er a powerful framework for extracting the two-factor structure of a set of observations, and are familiar in computational vision from several well-known lines of research. This paper shows how bilinear models can be used to learn the style-content structure of a pattern analysis or synthesis problem, which can then be generalized to solve related tasks using di#erentstyles and#or content. We focus on three kinds of tasks: extrapolating the style of data to unseen content classes, classifying data with known content under a novel style, and translating two sets of data, generated in di#erentstyles and with distinct content, into each other's styles. We show examples from color constancy, face pose estimation, shape-from-shading, typography and speech.
Citations
|
1770
|
Eigenfaces for recognition
– Turk, Pentland
- 1991
|
|
749
|
Visual Learning and Recognition of 3D Objects from Appearance
– Murase, Nayar
- 1995
|
|
642
|
Shape and motion from image streams under orthography: A factorization method
– Tomasi, Kanade
- 1992
|
|
372
|
Recognition by linear combinations of models
– UIIman, Basri
- 1989
|
|
202
|
What Is the Set of Images of an Object under All Possible Lighting Conditions
– Belhumeur, Kriegman
- 1998
|
|
168
|
Surface modeling with oriented particle systems
– Szeliski, Tonnesen
- 1992
|
|
102
|
Image representation for visual learning
– Beymer, Poggio
- 1996
|
|
101
|
Matrix Differential Calculus with Applications in Statistics and Econometrics, revised edition
– Magnus, Neudecker
- 1999
|
|
91
|
A low-dimensional representation of human faces for arbitrary lighting conditions
– Hallinan
|
|
82
|
Feature-based correspondence: an eigenvector approach
– Shapiro, Brady
- 1992
|
|
62
|
Redlich, “Statistical approach to shape from shading: Reconstruction of 3D face surfaces from single 2D images,” Neural Comput
– Atick, Griffin, et al.
- 1996
|
|
58
|
The measurement of highlights in color images
– Klinker, Shafer, et al.
- 1988
|
|
55
|
Method for computing the scene-illuminant chromaticity from specular highlights
– Lee
- 1986
|
|
49
|
Linear models of surface and illuminant spectra
– Marimont, Wandell
- 1992
|
|
44
|
Linear shape from shading
– Pentland
- 1990
|
|
23
|
Fluid Concepts and Creative Analogies
– Hofstadter
- 1995
|
|
18
|
The generic bilinear calibration-estimation problem
– Koenderink, Doorn
- 1997
|
|
16
|
Color constancy: surface color from changing illumination
– D’Zmura
- 1992
|
|
15
|
Family Discovery
– Omohundro
- 1996
|
|
14
|
Bayesian Decision Theory: the maximum local mass estimate
– Freeman, Brainard
- 1995
|
|
10
|
Connectionist generalizing for production: An example from GridFont
– Greber, Stork, et al.
- 1991
|
|
9
|
Modal Matching: A Method for Describing, Comparing, and Manipulating Digital Signals
– Sclaroff
- 1995
|
|
3
|
Separable mixture models: Separating style and content
– Tenenbaum, Freeman
- 1996
|
|
1
|
Geometry and photomnetry in 3D visual recognition
– Shashua
- 1992
|
