Modified the comportment of SpatialMatching with full_output = false

1 files changed, 14 insertions, 15 deletions

diff --git a/SpatialMatching.lua b/SpatialMatching.lua
index 10e0dc7..77632e3 100644
--- a/SpatialMatching.lua
+++ b/SpatialMatching.lua
@@ -1,33 +1,32 @@
 local SpatialMatching, parent = torch.class('nn.SpatialMatching', 'nn.Module')
 
 function SpatialMatching:__init(maxw, maxh, full_output)
+   -- If full_output is false, output is computed on elements of the first input
+   -- for which all the possible corresponding elements exist in the second input
+   -- In addition, if full_output is set to false, the pixel (1,1) of the first input
+   -- is supposed to correspond to the pixel (maxh/2, maxw/2) of the second one
+   -- If align_inputs is set, input[2] is assumed to be large enough
+   -- (that is input[1]:size(1) <= inputs[1]:size(1) - maxh + 1, same for w)
+   -- TODO full_output == true and align_inputs == false is probably useless
    parent.__init(self)
    self.maxw = maxw or 11
    self.maxh = maxh or 11
-   full_output = full_output or false
-   if full_output then
-      self.full_output = 1
-   else
-      self.full_output = 0
-   end
+   full_output = full_output or true
+   if full_output then self.full_output = 1 else self.full_output = 0 end
 end
 
 function SpatialMatching:updateOutput(input)
    -- input is a table of 2 inputs, each one being KxHxW
-   if self.full_output == 1 then
-      self.output:resize(input[1]:size(2), input[1]:size(3), self.maxh, self.maxw)
-   else
-      self.output:resize(input[1]:size(2)-self.maxh+1, input[1]:size(3)-self.maxw+1,
-			 self.maxh, self.maxw)
-   end
+   -- if not full_output, the 1st one is KxH1xW1 where H1 <= H-maxh+1, W1 <= W-maxw+1
+   self.output:resize(input[1]:size(2), input[1]:size(3), self.maxh, self.maxw)
    input[1].nn.SpatialMatching_updateOutput(self, input[1], input[2])
    return self.output
 end
 
 function SpatialMatching:updateGradInput(input, gradOutput)
-   -- todo this is probably wrong
-   self.gradInput1 = torch.Tensor():resizeAs(input[1]):zero()
-   self.gradInput2 = torch.Tensor():resizeAs(input[2]):zero()
+   -- TODO this is probably the wrong way
+   self.gradInput1 = torch.Tensor(input[1]:size()):zero()
+   self.gradInput2 = torch.Tensor(input[2]:size()):zero()
    input[1].nn.SpatialMatching_updateGradInput(self, input[1], input[2], gradOutput)
    self.gradInput = {self.gradInput1, self.gradInput2}
    return self.gradInput