Mercurial > ift6266
comparison writeup/techreport.tex @ 444:18841eeb433f
small syntax fix
| author | Xavier Glorot <glorotxa@iro.umontreal.ca> |
|---|---|
| date | Mon, 03 May 2010 18:43:24 -0400 |
| parents | 89a49dae6cf3 |
| children | b0622f78cfec |
comparison
equal
deleted
inserted
replaced
| 443:89a49dae6cf3 | 444:18841eeb433f |
|---|---|
| 167 | 167 |
| 168 \subsection{Pixel permutation} | 168 \subsection{Pixel permutation} |
| 169 | 169 |
| 170 This filter permuts neighbouring pixels. It selects first $\frac{complexity}{3}$ pixels randomly in the image. Each of them are then sequentially | 170 This filter permuts neighbouring pixels. It selects first $\frac{complexity}{3}$ pixels randomly in the image. Each of them are then sequentially |
| 171 exchanged to one other pixel in its $V4$ neighbourhood. Number of exchanges to the left, right, top, bottom are equal or does not differ from more than 1 | 171 exchanged to one other pixel in its $V4$ neighbourhood. Number of exchanges to the left, right, top, bottom are equal or does not differ from more than 1 |
| 172 if the number of selected pixel is not a multiple of 4. | 172 if the number of selected pixels is not a multiple of 4. |
| 173 | 173 |
| 174 It has has a probability of not being applied, at all, of 80\%. | 174 It has has a probability of not being applied, at all, of 80\%. |
| 175 | 175 |
| 176 | 176 |
| 177 \subsection{Distorsion gauss} | 177 \subsection{Distorsion gauss} |
| 204 $[12,12 + 20 \times complexity]$ and $[2,2 + 6 \times complexity]$. The result is normalized between $0$ and $1$. | 204 $[12,12 + 20 \times complexity]$ and $[2,2 + 6 \times complexity]$. The result is normalized between $0$ and $1$. |
| 205 We also create a symmetric averaging window, of the kernel size, with maximum value at the center. | 205 We also create a symmetric averaging window, of the kernel size, with maximum value at the center. |
| 206 For each image we sample uniformly from $3$ to $3 + 10 \times complexity$ pixels that will be averaging centers | 206 For each image we sample uniformly from $3$ to $3 + 10 \times complexity$ pixels that will be averaging centers |
| 207 between the original image and the filtered one. | 207 between the original image and the filtered one. |
| 208 We initialize to zero a mask matrix of the image size. For each selected pixel we add to the mask the averaging window centered to it. | 208 We initialize to zero a mask matrix of the image size. For each selected pixel we add to the mask the averaging window centered to it. |
| 209 The final image is computed from the following element-wise operation: $\frac{image + filtered_image \times mask}{mask+1}$. | 209 The final image is computed from the following element-wise operation: $\frac{image + filtered image \times mask}{mask+1}$. |
| 210 | 210 |
| 211 This filter has a probability of not being applied, at all, of 75\%. | 211 This filter has a probability of not being applied, at all, of 75\%. |
| 212 | 212 |
| 213 \subsection{"Ratures"} | 213 \subsection{"Ratures"} |
| 214 | 214 |