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 positionsto 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.