\relax 
\providecommand\hyper@newdestlabel[2]{}
\providecommand\HyperFirstAtBeginDocument{\AtBeginDocument}
\HyperFirstAtBeginDocument{\ifx\hyper@anchor\@undefined
\global\let\oldnewlabel\newlabel
\gdef\newlabel#1#2{\newlabelxx{#1}#2}
\gdef\newlabelxx#1#2#3#4#5#6{\oldnewlabel{#1}{{#2}{#3}}}
\AtEndDocument{\ifx\hyper@anchor\@undefined
\let\newlabel\oldnewlabel
\fi}
\fi}
\global\let\hyper@last\relax 
\gdef\HyperFirstAtBeginDocument#1{#1}
\providecommand\HyField@AuxAddToFields[1]{}
\providecommand\HyField@AuxAddToCoFields[2]{}
\@writefile{toc}{\contentsline {section}{\numberline {1}Introduction}{1}{section.1}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {2}Dataset}{1}{section.2}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {2.1}Data Augmentation}{1}{subsection.2.1}\protected@file@percent }
\newlabel{sec:augmentation}{{2.1}{1}{Data Augmentation}{subsection.2.1}{}}
\@writefile{lof}{\contentsline {figure}{\numberline {1}{\ignorespaces Sample images from the training dataset, showing the effect of data augmentation.\relax }}{2}{figure.1}\protected@file@percent }
\@writefile{toc}{\contentsline {section}{\numberline {3}Feature Extraction}{2}{section.3}\protected@file@percent }
\newlabel{sec:features}{{3}{2}{Feature Extraction}{section.3}{}}
\@writefile{toc}{\contentsline {section}{\numberline {4}Low-Level Features Models}{2}{section.4}\protected@file@percent }
\newlabel{sec:lowlevel}{{4}{2}{Low-Level Features Models}{section.4}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.1}Logistic Regression}{2}{subsection.4.1}\protected@file@percent }
\@writefile{lof}{\contentsline {figure}{\numberline {2}{\ignorespaces Logistic regression training curves for different low level features.\relax }}{3}{figure.2}\protected@file@percent }
\newlabel{fig:train_curves_color_logistic}{{2}{3}{Logistic regression training curves for different low level features.\relax }{figure.2}{}}
\@writefile{lof}{\contentsline {figure}{\numberline {3}{\ignorespaces Summary of training (semi-transparent) and validation (solid) performance for different low-level feature combinations with logistic regression.\relax }}{3}{figure.3}\protected@file@percent }
\newlabel{fig:color_recap_logistic}{{3}{3}{Summary of training (semi-transparent) and validation (solid) performance for different low-level feature combinations with logistic regression.\relax }{figure.3}{}}
\@writefile{lof}{\contentsline {figure}{\numberline {4}{\ignorespaces Visualization of the \emph  {Histogram of Oriented Gradients} for a sample image. It's evident how HOG features perfectly capture shape and edge information, crucial for road sign classification.\relax }}{3}{figure.4}\protected@file@percent }
\newlabel{fig:hogging}{{4}{3}{Visualization of the \emph {Histogram of Oriented Gradients} for a sample image. It's evident how HOG features perfectly capture shape and edge information, crucial for road sign classification.\relax }{figure.4}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.2}Feature reduction}{3}{subsection.4.2}\protected@file@percent }
\@writefile{toc}{\contentsline {subsection}{\numberline {4.3}Grayscale}{3}{subsection.4.3}\protected@file@percent }
\@writefile{lof}{\contentsline {figure}{\numberline {5}{\ignorespaces Summary of training (semi-transparent) and validation (solid) performance for different low-level feature combinations with logistic regression on grayscale images.\relax }}{4}{figure.5}\protected@file@percent }
\newlabel{fig:grayscale_recap_logistic}{{5}{4}{Summary of training (semi-transparent) and validation (solid) performance for different low-level feature combinations with logistic regression on grayscale images.\relax }{figure.5}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {4.4}Multi-Layer Perceptron}{4}{subsection.4.4}\protected@file@percent }
\@writefile{lof}{\contentsline {figure}{\numberline {6}{\ignorespaces Comparison of performance (both training and validation) on color and grayscale images with different MLPs.\relax }}{4}{figure.6}\protected@file@percent }
\newlabel{fig:recap_mlp}{{6}{4}{Comparison of performance (both training and validation) on color and grayscale images with different MLPs.\relax }{figure.6}{}}
\@writefile{toc}{\contentsline {section}{\numberline {5}Convolutional Neural Network}{4}{section.5}\protected@file@percent }
\newlabel{sec:cnn}{{5}{4}{Convolutional Neural Network}{section.5}{}}
\@writefile{lot}{\contentsline {table}{\numberline {1}{\ignorespaces CNN architecture. $C = 3$ for RGB, $C = 1$ for grayscale.\relax }}{5}{table.1}\protected@file@percent }
\newlabel{tab:cnn_arch}{{1}{5}{CNN architecture. $C = 3$ for RGB, $C = 1$ for grayscale.\relax }{table.1}{}}
\@writefile{lof}{\contentsline {figure}{\numberline {7}{\ignorespaces CNN training curves for RGB images.\relax }}{5}{figure.7}\protected@file@percent }
\newlabel{fig:cnn_training}{{7}{5}{CNN training curves for RGB images.\relax }{figure.7}{}}
\@writefile{toc}{\contentsline {subsection}{\numberline {5.1}Activation Maps}{5}{subsection.5.1}\protected@file@percent }
\@writefile{lof}{\contentsline {figure}{\numberline {8}{\ignorespaces Comparison of first-layer activation maps for RGB and grayscale inputs. The RGB maps show stronger responses to color features, while the grayscale maps focus more on edges, shapes and contrast.\relax }}{5}{figure.8}\protected@file@percent }
\newlabel{fig:activation_maps}{{8}{5}{Comparison of first-layer activation maps for RGB and grayscale inputs. The RGB maps show stronger responses to color features, while the grayscale maps focus more on edges, shapes and contrast.\relax }{figure.8}{}}
\@writefile{toc}{\contentsline {section}{\numberline {6}Conclusion}{5}{section.6}\protected@file@percent }
\gdef \@abspage@last{5}