361 |
that at some point, |
that at some point, |
362 |
the results from the different feature detectors, |
the results from the different feature detectors, |
363 |
such as local and global shapes and colors, |
such as local and global shapes and colors, |
364 |
combine to form an abstract \emph{feature vector} |
are combined to form an abstract \emph{feature vector} |
365 |
(see Fig.~\ref{fig-perceptual}). |
(see Fig.~\ref{fig-perceptual}). |
366 |
This feature vector is then used as an input to the |
The to compute which concept the particular |
|
networks which compute which concept the particular |
|
367 |
input corresponds to, in a simple perceptron-like |
input corresponds to, in a simple perceptron-like |
368 |
fashion\cite{rosenblatt62neurodynamics,widrow60adaptive}. |
fashion\cite{rosenblatt62neurodynamics,widrow60adaptive}. |
369 |
This is a configuration sometimes used in neural computation. |
This is a configuration sometimes used in neural computation. |
370 |
|
|
371 |
|
|
372 |
From this |
From this |
373 |
rough, qualitative model of visual perception |
rough, qualitative model |
374 |
we can formulate the following points: |
we can formulate the following points: |
375 |
To be distinguishable, \label{secfeaturevec} |
To be distinguishable, \label{secfeaturevec} |
376 |
\begin{itemize} |
\begin{itemize} |