this is for holding javascript data
Volker Strobel edited section_Texton_based_Machine_Learning__.tex
over 7 years ago
Commit id: 4c0059e3789e3f4eacebc4ee01626a6392e850e0
deletions | additions
diff --git a/section_Texton_based_Machine_Learning__.tex b/section_Texton_based_Machine_Learning__.tex
index 3e45ee0..d129552 100644
--- a/section_Texton_based_Machine_Learning__.tex
+++ b/section_Texton_based_Machine_Learning__.tex
...
\citet{varma2005statistical} originally introduced textons for
classifying different textures, showing that they outperform
computationally more complex algorithms, like Gabor
filters~\cite{varma2005statistical}. For the
histogram extraction classification, the approach compares texton histograms between a training set and the test sample and the class of the closest training sample is assigned to the test sampe. To extract histograms in the \emph{full sampling} setting, a small window---a kernel---is convolved over all image positions
to extract and
label patches. the frequency of textons is calculated.
Instead of convolving a kernel over the entire image, the kernel can be applied at randomly sampled image
positions, leading position\cite{sde2012sub}. %leading to similar texton histograms compared to the histograms when full sampling is used.
The choice of Modifying the number of samples allows for
modifying adjusting the computational effort, resulting in a
trade-off between accuracy and execution frequency. A disadvantage is
that it discards all information about the spatial arrangement of
textons---it does not make use of the \emph{Where} of the information,
just of the \emph{What}, which
might can result in different
areas images with
similar histograms.
\citet{de2009design} use textons as image features for distinguishing
...
variation, these objects should appear less varied. Using this method,
their MAV is successfully able to avoid obstacles in a $5m \times 5m$
office space and achieves an AUC of up to .97 on rather distorted
images.
Additionally, %Additionally, it performs better than or, at least, equal to
optical %optical flow estimations.