Wednesday, November 30, 2011
Prof. Stéphane Mallat - Classification with Stable and Informative Invariants
Description:
Image classification requires to reduce variability with invariant representations, which are stable to deformations and retain enough information for disrimination.Canonical invariants, Fourier or wavelet transforms do not satisfy these properties. Scattering operators iteratively apply wavelet transforms and modulus operators that remove the phase. They are proved to be invariant to translations and stable to deformations. Scattering operators also provide representations of stationary processes, including high order moments which discriminate different processes having same Fourier spectrum.
Classification results are shown on data bases of structured image patterns and textures.
Extension beyond the translation group, and group learning is discussed.
{Will appear in FYI on Nov 30, 2011
Start Time:
4:00 PM
End Time:
5:15 PM
Room:
2117
Venue:
*Note special location and time
Common Location Name:
Computer Science Instructional Center
Web Address:
www.cscamm.umd.edu/seminars/
Other Contact Information:
Debra Glapion
CSCAMM
+1 301 405 0652
dglapion@cscamm.umd.edu
www.cscamm.umd.edu
