\relax 
\@writefile{toc}{\contentsline {section}{\numberline {1}Part-based models for Object Recognition (50 points)}{1}}
\newlabel{fig:parts}{{1}{1}}
\@writefile{lof}{\contentsline {figure}{\numberline {1}{\ignorespaces \textbf  {(a)} An illustration of the parameters of an example deformable part-based model. Note that there is a connection parameter $c_{ij}$ corresponding to each edge $e_{ij}$, and an appearance parameter $u_i$ corresponding to each vertex $v_i$. \textbf  {(b)} Three training images showing the part locations and the last image depicting the learned structure model.}}{1}}
\citation{LiSuLimFeiFei}
\citation{LiSuXingFeiFei}
\@writefile{toc}{\contentsline {section}{\numberline {2}Scene Classification and Matching Using ObjectBank (50 points)}{5}}
\@writefile{lof}{\contentsline {figure}{\numberline {2}{\ignorespaces Diagram of computation pipeline for ObjectBank representation of an image. }}{5}}
\newlabel{fig:sail}{{2}{5}}
\citation{lazebnik06}
\citation{Felz08}
\citation{LiSuLimFeiFei}
\citation{LiSuXingFeiFei}
\@writefile{loa}{\contentsline {algorithm}{\numberline {1}{\ignorespaces Create ObjectBank Feature Vector}}{6}}
\newlabel{alg1}{{1}{6}}
\@writefile{lof}{\contentsline {figure}{\numberline {3}{\ignorespaces Overview of the Spatial Pyramid Kernel in which histograms of interest point (denoted by the three markers) are created for each subdivision in a 2-level pyramid. The method used in ObjectBank is very similar, except it only extracts the local max inside each subdivision. }}{6}}
\newlabel{fig:spm}{{3}{6}}
\@writefile{lof}{\contentsline {figure}{\numberline {4}{\ignorespaces Sample Test Images}}{7}}
\@writefile{lof}{\contentsline {subfigure}{\numberline{(a)}{\ignorespaces {}}}{7}}
\@writefile{lof}{\contentsline {subfigure}{\numberline{(b)}{\ignorespaces {}}}{7}}
\newlabel{fig:coke}{{4}{7}}
\bibstyle{plain}
\bibdata{Problem-Set-4-solutions}
\bibcite{Felz08}{1}
\bibcite{lazebnik06}{2}
\bibcite{LiSuLimFeiFei}{3}
\bibcite{LiSuXingFeiFei}{4}