Skip to content

Improve exercise1 #21

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
tomazas wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Improve exercise1 #21

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 21 additions & 9 deletions exercise1.m
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,9 @@
% -------------------------------------------------------------------------
% Part 1.1: Linear convolution
% -------------------------------------------------------------------------
clc
close all
clear all

% Read an example image
x = imread('peppers.png') ;
Expand All @@ -11,7 +14,8 @@
x = im2single(x) ;

% Visualize the input x
figure(1) ; clf ; imagesc(x) ;
figure(1) ;clf ; imagesc(x); title('original image');
pause

% Create a bank of linear filters
w = randn(5,5,3,10,'single') ;
Expand All @@ -20,15 +24,18 @@
y = vl_nnconv(x, w, []) ;

% Visualize the output y
figure(2) ; clf ; vl_imarraysc(y) ; colormap gray ;
figure(2) ; clf ; vl_imarraysc(y) ; colormap gray ; title('convolved image');
pause

% Try again, downsampling the output
y_ds = vl_nnconv(x, w, [], 'stride', 16) ;
figure(3) ; clf ; vl_imarraysc(y_ds) ; colormap gray ;
figure(3) ; clf ; vl_imarraysc(y_ds) ; colormap gray ; title('downsampled image');
pause

% Try padding
y_pad = vl_nnconv(x, w, [], 'pad', 4) ;
figure(4) ; clf ; vl_imarraysc(y_pad) ; colormap gray ;
figure(4) ; clf ; vl_imarraysc(y_pad) ; colormap gray ; title('padded image');
pause

% Manually design a filter
w = [0 1 0 ;
Expand All @@ -39,7 +46,7 @@
figure(5) ; clf ; colormap gray ;
subplot(1,2,1) ; imagesc(y_lap) ; title('filter output') ;
subplot(1,2,2) ; imagesc(-abs(y_lap)) ; title('- abs(filter output)') ;

pause

% -------------------------------------------------------------------------
% Part 1.2: Non-linear gating (ReLU)
Expand All @@ -51,15 +58,17 @@
z = vl_nnrelu(y) ;

figure(6) ; clf ; colormap gray ;
subplot(1,2,1) ; vl_imarraysc(y) ;
subplot(1,2,2) ; vl_imarraysc(z) ;
subplot(1,2,1) ; vl_imarraysc(y) ; title('convolved') ;
subplot(1,2,2) ; vl_imarraysc(z) ; title('relu') ;
pause

% -------------------------------------------------------------------------
% Part 1.2: Pooling
% -------------------------------------------------------------------------

y = vl_nnpool(x, 15) ;
figure(7) ; clf ; imagesc(y) ;
figure(7) ; clf ; imagesc(y) ; title('pooling') ;
pause

% -------------------------------------------------------------------------
% Part 1.3: Normalization
Expand All @@ -70,4 +79,7 @@
alpha = 1 ;
beta = 0.5 ;
y_nrm = vl_nnnormalize(x, [rho kappa alpha beta]) ;
figure(8) ; clf ; imagesc(y_nrm) ;
figure(8) ; clf ; imagesc(y_nrm) ; title('normalization') ;
pause
disp('done.');