diff options
| -rw-r--r-- | algorithm.tex | 6 | ||||
| -rw-r--r-- | experimental.tex | 52 | ||||
| -rw-r--r-- | kinect.tex | 10 | ||||
| -rw-r--r-- | uniqueness.tex | 2 |
4 files changed, 33 insertions, 37 deletions
diff --git a/algorithm.tex b/algorithm.tex index 3045929..c02a0e5 100644 --- a/algorithm.tex +++ b/algorithm.tex @@ -55,15 +55,15 @@ This suggests that all class conditionals $P(\bx | y)$ are multivariate normal and our generative model, although very simple, may be nearly optimal \cite{bishop06pattern}. -\begin{figure}[t] +\begin{figure*}[t] \centering - \includegraphics{graphics/limbs.pdf} + \includegraphics[width=0.99\textwidth]{graphics/limbs.pdf} \caption{Histograms of differences between 9 skeleton measurements $x_k$ (Section~\ref{sec:experiment}) and their expectation given the class $y$. In red, the p.d.f. of a normal distribution with mean and variance equal to the empirical mean and variance of the measurement} \label{fig:error marginals} -\end{figure} +\end{figure*} \subsection{Sequential hypothesis testing} diff --git a/experimental.tex b/experimental.tex index 07fd74b..90609da 100644 --- a/experimental.tex +++ b/experimental.tex @@ -49,7 +49,7 @@ detected figures, and the fitted skeleton of a person in a single frame. Skelet fit from roughly 1-5 meters away from the Kinect. For each frame with a skeleton we record the color image and the positions of the joints. -\begin{figure}[t] +\begin{figure*}[t] \begin{center} \includegraphics[width=0.99\textwidth]{graphics/hallway.png} \end{center} @@ -57,7 +57,7 @@ skeleton we record the color image and the positions of the joints. \caption{Experiment setting. Color image, depth image, and fitted skeleton as captured by the Kinect in a single frame} \label{fig:hallway} -\end{figure} +\end{figure*} For some frames, one or several joints are out of the frame or are occluded by another part of the body. In those cases, the coordinates of these joints are @@ -87,7 +87,8 @@ pairs of joints: %Shoulder-Elbow, Elbow-Wrist, ShoulderCenter-Spine, Spine-HipCenter, %HipCenter-HipSide, HipSide-Knee, Knee-Ankle. -\begin{table} +\vspace{\baselineskip} +\begin{table}[!h] \begin{center} \vspace{-1.5\baselineskip} \begin{tabular}{lll} @@ -98,6 +99,7 @@ Shoulder-Elbow & Spine-HipCenter & Knee-Ankle\\ \vspace{-2.5\baselineskip} \end{center} \end{table} +\vspace{1.5\baselineskip} Each detected skeleton also has an ID number obtained from the figure detection stage. When there are consecutive frames with the same ID, it means that figure @@ -142,7 +144,7 @@ happens if the noise from the Kinect is reduced. \begin{figure}[t] \begin{center} - \includegraphics[]{graphics/frames.pdf} + \includegraphics[width=0.49\textwidth]{graphics/frames.pdf} \end{center} \vspace{-1.5\baselineskip} \caption{Distribution of the frequency of each individual in the @@ -190,11 +192,11 @@ we reach 90\% accuracy at 60\% recall for a group size of 10 people. \begin{figure*}[t] \begin{center} \subfloat[Mixture of Gaussians]{ - \includegraphics[]{graphics/offline-nb.pdf} + \includegraphics[width=0.49\textwidth]{graphics/offline-nb.pdf} \label{fig:offline:nb} } \subfloat[Sequential Hypothesis Testing]{ - \includegraphics[]{graphics/offline-sht.pdf} + \includegraphics[width=0.49\textwidth]{graphics/offline-sht.pdf} \label{fig:offline:sht} } \caption{Results with 10-fold cross-validation for the top $n_p$ most present people} @@ -246,30 +248,22 @@ recognition rates mostly above 90\% for group sizes of 3 and 5. % \label{fig:online:nb} %} %\subfloat[Sequential hypothesis testing]{ -\parbox[t]{0.49\linewidth}{ \begin{center} \includegraphics[width=0.49\textwidth]{graphics/online-sht.pdf} \end{center} -} -\parbox[t]{0.49\linewidth}{ - \begin{center} - \includegraphics[width=0.49\textwidth]{graphics/face.pdf} - \end{center} -} -\end{figure} -\begin{figure} \vspace{-1.5\baselineskip} -\parbox[t]{0.48\linewidth}{ \caption{Results for the online setting, where $n_p$ is the size of the group as in Figure~\ref{fig:offline}} \label{fig:online} -} -\hspace{0.02\linewidth} -\parbox[t]{0.48\linewidth}{ +\end{figure} +\begin{figure}[t] + \begin{center} + \includegraphics[width=0.49\textwidth]{graphics/face.pdf} + \end{center} +\vspace{-1.5\baselineskip} \caption{Results for face recognition versus skeleton recognition with $n_p=5$ people} \label{fig:face} -} \end{figure} \subsection{Face recognition} @@ -300,28 +294,20 @@ less than 10\% at most thresholds. \begin{figure}[t] -\parbox[t]{0.49\linewidth}{ \begin{center} \includegraphics[width=0.49\textwidth]{graphics/back.pdf} \end{center} -} -\parbox[t]{0.49\linewidth}{ +\vspace{-1.5\baselineskip} +\caption{Results with people walking away from and toward the camera} +\label{fig:back} +\end{figure} +\begin{figure}[t] \begin{center} \includegraphics[width=0.49\textwidth]{graphics/var.pdf} \end{center} -} -\end{figure} -\begin{figure} \vspace{-1.5\baselineskip} -\parbox[t]{0.48\linewidth}{ -\caption{Results with people walking away from and toward the camera} -\label{fig:back} -} -\hspace{0.02\linewidth} -\parbox[t]{0.48\linewidth}{ \caption{Results with and without halving the variance of the noise} \label{fig:var} -} \end{figure} \subsection{Walking away} @@ -64,6 +64,16 @@ % \author{Anonymous ECCV submission} % \institute{Paper ID \ECCV12SubNumber} +\author{ \parbox{2 in}{\centering Jon Whiteaker\\ + {\tt\small jbw@berkeley.edu}} + \parbox{2 in}{ \centering Thibaut Horel\\ + {\tt\small thibaut.horel@ens.fr}} + \parbox{2 in}{ \centering Branislav Kveton\\ + {\tt\small branislav.kveton@technicolor.com}} +\\\\ +\centering\small Technicolor, Palo Alto, CA 94301, USA +} + \maketitle diff --git a/uniqueness.tex b/uniqueness.tex index 1814446..3d91dc8 100644 --- a/uniqueness.tex +++ b/uniqueness.tex @@ -68,7 +68,7 @@ output of the algorithm for the threshold $\delta$ is defined as: \begin{figure}[t] \begin{center} - \includegraphics[]{graphics/roc.pdf} + \includegraphics[width=0.49\textwidth]{graphics/roc.pdf} \end{center} \vspace{-1.5\baselineskip} \caption{ROC curve for several standard deviations of the noise and |