figure32 changed

discussion.tex

@@ -6,28 +6,49 @@ \subsection{Countermeasure}
 \begin{figure}[!t]
         \centering
         \subfigure{
-             \begin{minipage}[t]{0.11\textwidth}
+             \begin{minipage}[t]{0.13\textwidth}
                 \includegraphics[width=\textwidth]{fig/protection1.pdf}\\
                 \centering \footnotesize (a)
              \end{minipage}
         }
-        \hspace{0.3cm}
+        \hspace{0.1cm}
         \subfigure{
-             \begin{minipage}[t]{0.11\textwidth}
+             \begin{minipage}[t]{0.13\textwidth}
                 \includegraphics[width=\textwidth]{fig/protection2.pdf}\\
                 \centering \footnotesize (b)
              \end{minipage}
         }
-        \hspace{0.3cm}
+        \hspace{0.1cm}
         \subfigure{
-             \begin{minipage}[t]{0.11\textwidth}
+             \begin{minipage}[t]{0.13\textwidth}
                 \includegraphics[width=\textwidth]{fig/protection3.pdf}\\
                 \centering \footnotesize (c)
              \end{minipage}
         }
+        %\hspace{0.2cm}
+        \subfigure{
+             \begin{minipage}[t]{0.14\textwidth}
+                \includegraphics[width=\textwidth]{fig/protection1-1.pdf}\\
+                \centering \footnotesize (d)
+             \end{minipage}
+        }
+        %\hspace{0.1cm}
+        \subfigure{
+             \begin{minipage}[t]{0.14\textwidth}
+                \includegraphics[width=\textwidth]{fig/protection2-1.pdf}\\
+                \centering \footnotesize (e)
+             \end{minipage}
+        }
+        %\hspace{0.1cm}
+        \subfigure{
+             \begin{minipage}[t]{0.14\textwidth}
+                \includegraphics[width=\textwidth]{fig/protection3-1.pdf}\\
+                \centering \footnotesize (f)
+             \end{minipage}
+        }
         \caption{Examples of our new pattern lock mechanism. A true pattern of solid lines is shown
-        in (a). Our first change allows the user to skip some dots when creating a pattern. 
-        For example, in (a) the central dot in the first line is skipped. 
+        in (a). Our first change allows the user to skip some dots when creating a pattern.
+        For example, in (a) the central dot in the first line is skipped.
         Our second change requires the user to draw some random pattern structures (e.g. the dash lines in c and d)  before or after drawing the true pattern.
         The first change makes it difficult for the tracking algorithm to identify which dots are skipped, and the second change
          forces the attacker to use multiple video recordings to identify the true pattern. As a result, this new pattern lock mechanism decreases the
@@ -83,7 +104,7 @@ \subsection{A Possible Remedy}
 We propose a countermeasure to make it difficult to obtain a meaningful video footage.
 When design the approach, we try to find a balance between the usability and the security.
 Our approach requires making two  changes to the pattern lock: (1) when forming
-a pattern the user can skip some of the dots in a vertical, horizontal, or diagonal line (Figure~\ref{fig:protection} a); (2) 
+a pattern the user can skip some of the dots in a vertical, horizontal, or diagonal line (Figure~\ref{fig:protection} a); (2)
 before or after drawing the correct pattern, the user is asked to draw a random pattern to confuse the attacker (Figure~\ref{fig:protection} b and c).
 For the second change, we relax the rules for creating a pattern to allow a touching dot to be visited multiple times. Further, for
 purpose of increasing the number of candidate patterns, we also change the rule that a previously unvisited dot can be

experiment_setup.tex

@@ -46,7 +46,7 @@ \section{Experimental Setup \label{sec:setup}}
     total length of all line segments that form the pattern (see
     Figure~\ref{fig:intersection-overlap}~a), $I_{P}$ are the number of
     intersections (which are also termed as "knight moves" in some prior
-    work~\cite{vonZezschwitz:2015:EDB:2702123.2702202}, see
+    work~\cite{Von2015Easy}, see
     Figure~\ref{fig:intersection-overlap}~b) and $O_{P}$ are the number of
     overlapping linear segments (see
     Figure~\ref{fig:intersection-overlap}~c).  To calculate the line length,
@@ -106,7 +106,7 @@ \section{Experimental Setup \label{sec:setup}}
 
 
     \subsection{Video Recording and Preprocessing}
-    
+
     \noindent\textbf{User Participation} We recruited ten postgraduate students (five male and five female
     students) from Northwest University to reproduce the 120 patterns (collected from users)
     and the 60 most complex patterns (see Section~\ref{sec:overall_rate})  on three target mobile phones:

intro.tex

@@ -181,7 +181,7 @@ \section{Introduction\label{sec:intro}}
 operating system (OS) allows five tries before
 locking the device, our attack represents a real threat for pattern lock.
 
-\item \emph{Identifying New Vulnerabilities.} According to a recent study~\cite{DBLP:conf/soups/2014}, direct
+\item \emph{Identifying New Vulnerabilities.} According to a recent study~\cite{Harbach2014It}, direct
 observation techniques, e.g. shoulder surfing, are considered to be a low risk due to the close distance between the
 attacker and the user (in order to gain a clear sight of the device screen). As a result, many users may underestimate
 the risks of using pattern lock in public places. Under our attack, filming can be carried out at a distance of 2

preamble.tex

@@ -10,7 +10,8 @@
 \usepackage{float}
 \usepackage{algorithm}
 \usepackage{algorithmic}
-\usepackage[colorlinks, linkcolor=black, anchorcolor=black, citecolor=blue]{hyperref}
+%\usepackage[colorlinks, linkcolor=black, anchorcolor=black, citecolor=blue]{hyperref}
+\usepackage[colorlinks, linkcolor=black, anchorcolor=black, citecolor=blue, urlcolor=black]{hyperref}
 \usepackage{cite}
 \usepackage{balance}
 \usepackage{color}
@@ -40,4 +41,4 @@
     {\node[shape = circle, draw, inner sep = 1pt]
     (char) {\phantom{\ifblank{#1}{#2}{#1}}};%
     \node at (char.center) {\makebox[0pt][c]{#2}};}}
-\robustify{\circled}
+\robustify{\circled}