The IPCA-ICA method based on incremental update of
the non-Gaussian independent vectors has been introduced.
The method concentrates on a challenging issue of computing
dominating non-Gaussian vectors from an incrementally
arriving high-dimensional data stream without computing
the corresponding covariance matrix and without knowing
the data in advance.
It is very efficient inmemory usage (only one input image
is needed at every step) and it is very efficient in the calculation
of the first basis vectors (unwanted vectors do not need
to be calculated). In addition to these advantages, this algorithm
gives an acceptable recognition success rate in comparison
with the PCA and the LDA algorithms.