tiny

yeguixin · Apr 27, 2017 · 4c06006 · 4c06006
1 parent c4cde2e
commit 4c06006
Show file tree

Hide file tree

Showing 6 changed files with 16 additions and 15 deletions.
diff --git a/TIFS.pdf b/TIFS.pdf
diff --git a/TIFS.tex b/TIFS.tex
@@ -18,23 +18,23 @@
             The Network and Distributed System Security Symposium (NDSS) 2017~\cite{ye2017cracking}.
             The extended version makes the following several additional contributions over the conference
             paper, providing new insights to the original paper:
-            
-            
+
+
             (1)	It quantifies the strength of locking patterns using the guessing probability, showing that traditional approaches for quantify the pattern lock security is unreliable under video-side channel attacks (section~\ref{section: pattern-complexity-classification});
-            
-            
+
+
             (2)	It provides new evaluations to understand the impact of the screen size and camera on the success rate of the attack (section~\ref{section: screen-size and cameras});
-            
-            
+
+
             (3)	It extends the attacking method to break PIN-based passwords, demonstrating the applicability of the attack on PIN-based passwords (section~\ref{section: attacking-pin-passwords});
-            
-            
+
+
             (4)	It includes a limited study to evaluate the effectiveness of the attack, where the video footage only captures he fingertip (section~\ref{section: limited-study});
-            
-            
+
+
             (5)	It proposes a simple, yet effective countermeasure. By making some simple modifications to the way a pattern lock is generated, the success rate of the attack will drop significantly (section~\ref{section: potential-remedy});
-           
-           
+
+
             (6)	It extends the related work section to discuss some of the recent studies on the security of Android pattern lock (section~\ref{section: relate-work}).
 
     }

diff --git a/details.tex b/details.tex
@@ -176,7 +176,7 @@ \subsection{Video preprocessing \label{sec:identify}}
 In practice, an experienced attacker would wait patiently to avoid this
 complicated situation by finding the right time for filming (e.g. for a screen
 lock, the time is just after the device is retrieved).
-The attacker could also watch the video to manually cut it to ensure \FIXED{to} obtain the correct video segment.
+The attacker could also watch the video to manually cut it to ensure to obtain the correct video segment.
 It is
 worthwhile to mention that automatically identifying the pattern unlocking process is
 not central to our attack because an experienced attacker can watch the video to manually cut it to ensure the tracking algorithm (described in the section) receives a quality input.
@@ -301,7 +301,7 @@ \subsection{Track fingertip locations}
         alike the correct pattern.
 
         \subsubsection{Noisy Points Calibration}
-            During tracking process, the TLD algorithm may report mistaken position of a tracked object as rapid deformation of the tracked object. This can affect the shape of fingertip movement, leading to extract \FIXED{incorrect} geometric information of fingertip movement.
+            During tracking process, the TLD algorithm may report mistaken position of a tracked object as rapid deformation of the tracked object. This can affect the shape of fingertip movement, leading to extract incorrect geometric information of fingertip movement.
 
 \subsection{Filming angle transformation}
 \label{sec:transformation}

diff --git a/intro.tex b/intro.tex
@@ -147,6 +147,7 @@ \section{Introduction\label{sec:intro}}
         }
         \caption{Examples of scenarios in which a mobile phone camera is used to film the unlocking process.
         In these scenarios, the camera does not need to have a clear sight of the screen.}
+        \FIXED{Change the pictures of scenario.}
         \label{fig:fig1}
     \end{figure}
 

diff --git a/ref/07336560.pdf b/ref/07336560.pdf
diff --git a/related.tex b/related.tex
@@ -55,7 +55,7 @@ \section{Related Work}
 Aviv \emph{et al.} developed a brutal-force based algorithm to
 crack pattern lock~\cite{Aviv2015Is}. Their results show that patterns
 generated on a grid of $3\times3$ and $4\times4$ dots can be cracked
-within thousands of guesses, where simple patterns \FIXED{need} less
+within thousands of guesses, where simple patterns need less
 attempts. In more recent work~\cite{Aviv2016Anlyzing}, they study the representativeness of
 pattern locks collected through various methods. Their work suggests
 that there are subtle differences for patterns collected using a pen-and-paper