avant complexite

[canny.git] / experiments.tex

diff --git a/experiments.tex b/experiments.tex
index aa8a3e94dd8b4c872185249d20ffcedeb6fd4c44..8194a5398fca17be6938580d63e9785f81b8857b 100644 (file)
--- a/experiments.tex
+++ b/experiments.tex
@@ -1,22 +1,61 @@
-For whole experiments, the whole set of 10000 images 
+For all the experiments, the whole set of 10,000 images 

 of the BOSS contest~\cite{Boss10} database is taken. 
 In this set, each cover is a $512\times 512$
 grayscale digital image in a RAW format. 
 We restrict experiments to 
 this set of cover images since this paper is more focused on 
 of the BOSS contest~\cite{Boss10} database is taken. 
 In this set, each cover is a $512\times 512$
 grayscale digital image in a RAW format. 
 We restrict experiments to 
 this set of cover images since this paper is more focused on 

-the methodology than benchmarking.    
+the methodology than on benchmarks.    
+
+We use the matrices $\hat{H}$ 
+generated by the integers given
+in Table~\ref{table:matrices:H}
+as introduced in~\cite{FillerJF11}, since these ones have experimentally 
+be proven to have the best modification efficiency.
+For instance if the rate between the size of the message and the size of the 
+cover vector
+is 1/4, each number in $\{81, 95, 107, 121\}$ is translated into a binary number 
+and each one constitutes thus a column of $\hat{H}$. 
+
+\begin{table}
+$$
+\begin{array}{|l|l|}
+\hline
+\textrm{Rate} & \textrm{Matrix generators} \\
+\hline
+{1}/{2} & \{71,109\}\\
+\hline
+{1}/{3} & \{95, 101, 121\}\\
+\hline
+{1}/{4} & \{81, 95, 107, 121\}\\
+\hline
+{1}/{5} & \{75, 95, 97, 105, 117\}\\
+\hline
+{1}/{6} & \{73, 83, 95, 103, 109, 123\}\\
+\hline
+{1}/{7} & \{69, 77, 93, 107, 111, 115, 121\}\\
+\hline
+{1}/{8} & \{69, 79, 81, 89, 93, 99, 107, 119\}\\
+\hline
+{1}/{9} & \{69, 79, 81, 89, 93, 99, 107, 119, 125\}\\
+\hline
+\end{array}
+$$
+\caption{Matrix Generator for $\hat{H}$ in STC}\label{table:matrices:H}
+\end{table}
+
+

 Our approach is always compared to Hugo~\cite{DBLP:conf/ih/PevnyFB10}
 and to EAISLSBMR~\cite{Luo:2010:EAI:1824719.1824720}.
 The former is the least detectable information hiding tool in spatial domain 
 Our approach is always compared to Hugo~\cite{DBLP:conf/ih/PevnyFB10}
 and to EAISLSBMR~\cite{Luo:2010:EAI:1824719.1824720}.
 The former is the least detectable information hiding tool in spatial domain 

-and the later is the work that is close to ours, as far as we know. 
+and the latter is the work that is the closest to ours, as far as we know. 

 
 
 
 First of all,  in our experiments and with the adaptive scheme, 
 the average size of the message that can be embedded is 16,445 bits.
 
 
 
 First of all,  in our experiments and with the adaptive scheme, 
 the average size of the message that can be embedded is 16,445 bits.

-Its corresponds to an  average payload of 6.35\%. 
+It corresponds to an  average payload of 6.35\%. 

 The two other tools will then be compared with this payload. 
 The two other tools will then be compared with this payload. 

-The Sections~\ref{sub:quality} and~\ref{sub:steg} respectively present 
+Sections~\ref{sub:quality} and~\ref{sub:steg} respectively present 

 the quality analysis and the security of our scheme. 
 
 
 the quality analysis and the security of our scheme. 
 
 


@@ -79,7 +118,7 @@ HUGO and STABYLO with  STC+adaptive parameters.
 Results are summarized in Table~\ref{table:quality}.
 Let us give an interpretation of these experiments.
 First of all, the adaptive strategy produces images with lower distortion 
 Results are summarized in Table~\ref{table:quality}.
 Let us give an interpretation of these experiments.
 First of all, the adaptive strategy produces images with lower distortion 

-than the one of images resulting from the 10\% fixed strategy.
+than the images resulting from the 10\% fixed strategy.

 Numerical results are indeed always greater for the former strategy than 
 for the latter one.
 These results are not surprising since the adaptive strategy aims at 
 Numerical results are indeed always greater for the former strategy than 
 for the latter one.
 These results are not surprising since the adaptive strategy aims at 


@@ -93,16 +132,16 @@ the two least significant bits.
 
 If we combine \emph{adaptive} and \emph{STC} strategies 
 (which leads to an average embedding rate equal to 6.35\%)
 
 If we combine \emph{adaptive} and \emph{STC} strategies 
 (which leads to an average embedding rate equal to 6.35\%)

-our approach  provides equivalent metrics than HUGO.
+our approach  provides metrics equivalent to those provided by HUGO.

 In this column STC(7) stands for embedding data in the LSB whereas
 In this column STC(7) stands for embedding data in the LSB whereas

-in STC(6), data are hidden in the two last significant bits. 
+in STC(6), data are hidden in the last two  significant bits. 

 
 
 
 The quality variance between HUGO and STABYLO for these parameters
 is given in bold font. It is always close to 1\% which confirms 
 the objective presented in the motivations:
 
 
 
 The quality variance between HUGO and STABYLO for these parameters
 is given in bold font. It is always close to 1\% which confirms 
 the objective presented in the motivations:

-providing an efficient steganography approach with a lightweight manner.
+providing an efficient steganography approach in a lightweight manner.

 
 
 Let us now compare the STABYLO approach with other edge based steganography
 
 
 Let us now compare the STABYLO approach with other edge based steganography


@@ -118,20 +157,22 @@ give quality metrics for fixed embedding rates from a large base of images.
 
 
 
 
 
 

-The steganalysis quality of our approach has been evaluated through the two 
-AUMP~\cite{Fillatre:2012:ASL:2333143.2333587}
-and Ensemble Classifier~\cite{DBLP:journals/tifs/KodovskyFH12} based steganalysers.
-Both aim at detecting hidden bits in grayscale natural images and are 
-considered as the state of the art of steganalysers in spatial domain~\cite{FK12}.
-The former approach is based on a simplified parametric model of natural images.
-Parameters are firstly estimated and an adaptive Asymptotically Uniformly Most Powerful
-(AUMP) test is designed (theoretically and practically), to check whether
-an image has stego content or not.
-This approach is dedicated to verify whether LSB has been modified or not.
-In the latter, the authors show that the 
-machine learning step, which is often
-implemented as a support vector machine,
-can be favorably executed thanks to an ensemble classifier.
+The steganalysis quality of our approach has been evaluated through the % two 
+% AUMP~\cite{Fillatre:2012:ASL:2333143.2333587}
+% and
+Ensemble Classifier~\cite{DBLP:journals/tifs/KodovskyFH12} based steganalyser.
+This approach  aims at detecting hidden bits in grayscale natural
+images and is 
+considered as state of the art steganalysers in the spatial domain~\cite{FK12}.
+%The former approach is based on a simplified parametric model of natural images.
+% Parameters are firstly estimated and an adaptive Asymptotically Uniformly Most Powerful
+% (AUMP) test is designed (theoretically and practically), to check whether
+% an image has stego content or not.
+% This approach is dedicated to verify whether LSB has been modified or not.
+% , the authors show that the 
+% machine learning step, which is often
+% implemented as a support vector machine,
+% can be favorably executed thanks to an ensemble classifier.

 
 
 \begin{table*}
 
 
 \begin{table*}


@@ -145,8 +186,8 @@ Embedding & Fixed &   \multicolumn{3}{|c|}{Adaptive (about 6.35\%)}  & \multicol
 \hline
 Rate & 10\% &  + sample &   +STC(7) & +STC(6)   & 10\%& 6.35\%& 10\%& 6.35\%\\ 
 \hline
 \hline
 Rate & 10\% &  + sample &   +STC(7) & +STC(6)   & 10\%& 6.35\%& 10\%& 6.35\%\\ 
 \hline

-AUMP & 0.22 & 0.33 & 0.39  &   0.45    &  0.50 &  0.50 & 0.49 & 0.50 \\
-\hline
+%AUMP & 0.22 & 0.33 & 0.39  &   0.45    &  0.50 &  0.50 & 0.49 & 0.50 \\
+%\hline

 Ensemble Classifier & 0.35 & 0.44 & 0.47 & 0.47     & 0.48 &  0.49  &  0.43  & 0.46 \\
 
 \hline
 Ensemble Classifier & 0.35 & 0.44 & 0.47 & 0.47     & 0.48 &  0.49  &  0.43  & 0.46 \\
 
 \hline


@@ -158,14 +199,19 @@ Ensemble Classifier & 0.35 & 0.44 & 0.47 & 0.47     & 0.48 &  0.49  &  0.43  & 0
 
 
 Results are summarized in Table~\ref{table:steganalyse}.
 
 
 Results are summarized in Table~\ref{table:steganalyse}.

-First of all, STC outperforms the sample strategy for the two steganalysers, as 
+First of all, STC outperforms the sample strategy %for % the two steganalysers
+ as 

 already noticed in the quality analysis presented in the previous section. 
 Next, our approach is more easily detectable than HUGO, which
 is the most secure steganographic tool, as far as we know. 
 However by combining \emph{adaptive} and \emph{STC} strategies
 already noticed in the quality analysis presented in the previous section. 
 Next, our approach is more easily detectable than HUGO, which
 is the most secure steganographic tool, as far as we know. 
 However by combining \emph{adaptive} and \emph{STC} strategies

-our approach obtains similar results than HUGO ones.
+our approach obtains similar results to HUGO ones.
+
+%%%%et pour b= 6 ?

 
 

+
+Compared to EAILSBMR, we obtain better results when the strategy is 
+\emph{adaptive}. 

 However due to its 
 However due to its 

-huge number of features integration, it is not lightweight, which justifies 
+huge number of integration features, it is not lightweight, which justifies 

 in the authors' opinion the consideration of the proposed method.   
 in the authors' opinion the consideration of the proposed method.   

-