Next: Wavelet basis and Wavelet
Up: Introduction to Wavelet Networks
Previous: Introduction to Wavelet Networks
Direct Calculation of Weights
Wavelet functions are not necessarily orthogonal. For a given family
of wavelets it is therefore not generally possible
to calculate a wavelet coefficient directly by a simple projection of the
wavelet
onto the considered function. In
[Daubechies, 1990,Daugman, 1988] it was therefore proposed to use
eq. (2) to find the optimal coefficients for each
fixed wavelet. Because optimization is a slow process, we want to suggest a
direct calculation for the case of a finite wavelet family.
The correct coeffs.
are computed by projecting the dual wavelets
.
The wavelet
is the dual wavelet to the
wavelet
if
|
(4) |
With
,
we can write
|
(5) |
and we find
to be
|
(6) |
where
.
Given a family of optimized wavelets of a WN for the function
, we can compute the orthogonal projection of a function
into the subspace
(see
(4)), i.e.
with |
(7) |
The method to compute the orthogonal projection of a function into
the subspace
is mathematically equivalent to using the
pseudo-inverse of directly. However, using the dual wavelets
of eq. (7) will proof to be computationally more
efficient: For our tracking experiment we will have to deform the
entire WN affinely,
which means that the pseudo-inverse has to be recomputed. The matrix
, on the other hand, is invariant, except for a factor,
to affine deformations of the WN, and only the projections
need to be recomputed.
Next: Wavelet basis and Wavelet
Up: Introduction to Wavelet Networks
Previous: Introduction to Wavelet Networks
Volker Krueger
2001-05-31