AI : neural net try : image recognision

[bluatigro]

this is why i needed a getpixel()

WARNING :
gethdc not present jet
so this does not work jet

the net trys to seperate square's and circle's


'' bluatigro 14 nov 2018
'' ann bmp
'' based on :
''http://code.activestate.com/recipes/578148-simple-back-propagation-neural-network-in-python-s/
global ni , nh , no , size
size = 63
ni = size * size
nh = size * sqr( size )
no = 1
dim ai( ni ) , ah( nh )
dim ao( no ) , wish( no )
dim wi( ni , nh )
dim wo( nh , no )
dim ci( ni , nh )
dim co( nh , no )
dim od( nh ) , hd( nh )
open "ann bmp" for graphics as #m
 #m "trapclose [quit]"
 
 call init
 ''init inpout and output paterns


 ''let NN live and learn
 for e = 0 to 1000
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   ''fil target
   wish( 0 ) = dice
   call calc
   fout = backprop( .5 , .5 )
   #m "goto 0 " ; size + 50
   #m "down"
   #m "\live : " ; e ; " | error : " ; fout
   #m "up"
 next e
 notice "Lerning ready ."
 fout = 0
 for e = 0 to 100
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   call calc
   if dice > 0 then
     if ao( 0 ) < 0 then
       fout = fout + 1
     end if
   else
     if ao( 0 ) > 0 then
       fout = fout + 1
     end if
   end if
 next e
 #m "goto 0 " ; size + 50
 #m "down"
 #m "\error : " ; fout
 #m "up"
 notice "[ game over ]"
wait
[quit]
 close #m
end
function getpixel( hDC , x , y )
 calldll #gdi32 , "GetPixel" _
 , hDC as ulong _
 , x as long _
 , y as long _
 , getpixel as ulong
end function
function in( x , y )
 in = x + y * size
end function
function drawsquare()
 r = irange( 10 , 30 )
 x = irange( r + 5 , 63 - r - 5 )
 y = irange( r + 5 , 63 - r - 5 )
 if irange( 0 , 1 ) then
   uit = 1
   #m "goto " ; x - r ; " " ; y - r
   #m "down"
   #m "boxfilled " ; x + r ; " " ; y + r
   #m "up"
 else
   uit = -1
   #m "goto " ; x ; " " ; y
   #m "down"
   #m "circlefilled " ; r
   #m "up"
 end if
 drawsquare = uit
end function
function irange( l , h )
 irange = int( range( l , h + 1 ) )
end function
function range( l , h )
 range = rnd(0) * ( h - l ) + l
end function
sub init
''init neural net
 for i = 0 to ni
   ai( i ) = 1
 next i
 for i = 0 to no - 1
   ao( i ) = 1
 next i
 for i = 0 to ni
   for h = 0 to nh - 1
     wi( i , h ) = range( -1 , 1 )
   next h
 next i
 for h = 0 to nh - 1
   for o = 0 to no - 1
     wo( h , o ) = range( -1 , 1 )
   next o
 next h
end sub
sub calc
''forwart pass of neural net
 for h = 0 to nh
   sum = 0
   for i = 0 to ni
     sum = sum + ai( i ) * wi( i , h )
   next i
   ah( h ) = signoid( sum / ni )
 next h
 for o = 0 to no
   sum = 0
   for h = 0 to nh
     sum = sum + ah( h ) * wo( h , o )
   next h
   ao( o ) = signoid( sum / nh )
 next o
end sub
function tanh( x )
 tanh = ( 1 - exp( -2 * x ) ) _
 / ( 1 + exp( -2 *x ) )
end function
function signoid( x )
 signoid = tanh( x )
end function
function dsignoid( x )
 dsignoid = 1 - x ^ 2
end function
function backprop( n , m )
'' http://www.youtube.com/watch?v=aVId8KMsdUU&feature=BFa&list=LLldMCkmXl4j9_v0HeKdNcRA
'' calc output deltas
'' we want to find the instantaneous rate of change of ( error with respect to weight from node j to node k)
'' output_delta is defined as an attribute of each ouput node. It is not the final rate we need.
'' To get the final rate we must multiply the delta by the activation of the hidden layer node in question.
'' This multiplication is done according to the chain rule as we are taking the derivative of the activation function
'' of the ouput node.
'' dE/dw[j][k] = (t[k] - ao[k]) * s'( SUM( w[j][k]*ah[j] ) ) * ah[j]
 for k = 0 to no - 1
   fout = wish( k ) - ao( k )
   od( k ) = fout * dsignoid( ao( k ) )
 next k
'' update output weights
 for j = 0 to nh
   for k = 0 to no
'' output_deltas[k] * self.ah[j]
'' is the full derivative of
'' dError/dweight[j][k]
     c = od( k ) * ah( nh , j )
     wo( j , k ) = wo( j , k ) + n * c + m * co( j , k )
     co( j , k ) = c
   next k
 next j
'' calc hidden deltas input layer
 for j = 0 to nh
   fout = 0
   for k = 0 to no
     fout = fout + od( k ) * wo( j , k )
   next k
   hd( j ) = fout * dsignoid( ao( j ) )
 next j
'' update input weights
 for i = 0 to ni
   for j = 0 to nh
     c = hd( j ) * ai( i )
     wi( i , j ) = wi( i , j ) _
     + n * c + m * ci( i , j )
     ci( i , j ) = c
   next j
 next i
 fout = 0
 for k = 0 to no
   fout = fout _
   + ( wish( k ) - ao( k ) ) ^ 2
 next k
 backprop = fout / 2
end function


[bluatigro]

update :
all nesery stuf is put in

it shoot work now


'' bluatigro 16 nov 2018
'' ann bmp
'' based on :
''http://code.activestate.com/recipes/578148-simple-back-propagation-neural-network-in-python-s/
global ni , nh , no , size , hw , hdc
size = 63
ni = size * size
nh = size * sqr( size )
no = 1
dim ai( ni ) , ah( nh )
dim ao( no ) , wish( no )
dim wi( ni , nh )
dim wo( nh , no )
dim ci( ni , nh )
dim co( nh , no )
dim od( nh ) , hd( nh )
open "ann bmp" for graphics as #m
 #m "trapclose [quit]"

 call init

 hw = hwnd( #m )
 calldll #user32, "GetDC" _
 , hw as ulong _   'handle of window or graphicbox client area
 , hdc as ulong    'returns handle of Device Context - 0=failure

 ''let NN live and learn
 for e = 0 to 1000
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   ''fil target
   wish( 0 ) = dice
   call calc
   fout = backprop( .5 , .5 )
   #m "goto 0 " ; size + 50
   #m "down"
   #m "\live : " ; e ; " | error : " ; fout
   #m "up"
 next e
 notice "Lerning ready ."
 fout = 0
 for e = 0 to 100
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   call calc
   if dice > 0 then
     if ao( 0 ) <= 0 then
       fout = fout + 1
     end if
   else
     if ao( 0 ) > 0 then
       fout = fout + 1
     end if
   end if
 next e
 #m "goto 0 " ; size + 50
 #m "down"
 #m "\error : " ; fout
 #m "up"
 notice "[ game over ]"
wait
[quit]
 calldll #user32 , "ReleaseDC" _
 , hw as ulong _    'window handle
 , hdc as ulong _  'device context
 , ret as long     'nonzero=success
 close #m
end
function getpixel( hDC , x , y )
 calldll #gdi32 , "GetPixel" _
 , hDC as ulong _
 , x as long _
 , y as long _
 , getpixel as ulong
end function
function in( x , y )
 in = x + y * size
end function
function drawsquare()
 r = irange( 10 , 30 )
 x = irange( r + 5 , 63 - r - 5 )
 y = irange( r + 5 , 63 - r - 5 )
 if irange( 0 , 1 ) then
   uit = 1
   #m "goto " ; x - r ; " " ; y - r
   #m "down"
   #m "boxfilled " ; x + r ; " " ; y + r
   #m "up"
 else
   uit = -1
   #m "goto " ; x ; " " ; y
   #m "down"
   #m "circlefilled " ; r
   #m "up"
 end if
 drawsquare = uit
end function
function irange( l , h )
 irange = int( range( l , h + 1 ) )
end function
function range( l , h )
 range = rnd(0) * ( h - l ) + l
end function
sub init
''init neural net
 for i = 0 to ni
   ai( i ) = 1
 next i
 for i = 0 to no - 1
   ao( i ) = 1
 next i
 for i = 0 to ni
   for h = 0 to nh - 1
     wi( i , h ) = range( -1 , 1 )
   next h
 next i
 for h = 0 to nh - 1
   for o = 0 to no - 1
     wo( h , o ) = range( -1 , 1 )
   next o
 next h
end sub
sub calc
''forwart pass of neural net
 for h = 0 to nh
   sum = 0
   for i = 0 to ni
     sum = sum + ai( i ) * wi( i , h )
   next i
   ah( h ) = signoid( sum / ni )
 next h
 for o = 0 to no
   sum = 0
   for h = 0 to nh
     sum = sum + ah( h ) * wo( h , o )
   next h
   ao( o ) = signoid( sum / nh )
 next o
end sub
function tanh( x )
 tanh = ( 1 - exp( -2 * x ) ) _
 / ( 1 + exp( -2 *x ) )
end function
function signoid( x )
 signoid = tanh( x )
end function
function dsignoid( x )
 dsignoid = 1 - x ^ 2
end function
function backprop( n , m )
'' http://www.youtube.com/watch?v=aVId8KMsdUU&feature=BFa&list=LLldMCkmXl4j9_v0HeKdNcRA
'' calc output deltas
'' we want to find the instantaneous rate of change of ( error with respect to weight from node j to node k)
'' output_delta is defined as an attribute of each ouput node. It is not the final rate we need.
'' To get the final rate we must multiply the delta by the activation of the hidden layer node in question.
'' This multiplication is done according to the chain rule as we are taking the derivative of the activation function
'' of the ouput node.
'' dE/dw[j][k] = (t[k] - ao[k]) * s'( SUM( w[j][k]*ah[j] ) ) * ah[j]
 for k = 0 to no - 1
   fout = wish( k ) - ao( k )
   od( k ) = fout * dsignoid( ao( k ) )
 next k
'' update output weights
 for j = 0 to nh
   for k = 0 to no
'' output_deltas[k] * self.ah[j]
'' is the full derivative of
'' dError/dweight[j][k]
     c = od( k ) * ah( nh , j )
     wo( j , k ) = wo( j , k ) + n * c + m * co( j , k )
     co( j , k ) = c
   next k
 next j
'' calc hidden deltas input layer
 for j = 0 to nh
   fout = 0
   for k = 0 to no
     fout = fout + od( k ) * wo( j , k )
   next k
   hd( j ) = fout * dsignoid( ao( j ) )
 next j
'' update input weights
 for i = 0 to ni
   for j = 0 to nh
     c = hd( j ) * ai( i )
     wi( i , j ) = wi( i , j ) _
     + n * c + m * ci( i , j )
     ci( i , j ) = c
   next j
 next i
 fout = 0
 for k = 0 to no
   fout = fout _
   + ( wish( k ) - ao( k ) ) ^ 2
 next k
 backprop = fout / 2
end function


[tenochtitlanuk]

Array ah( ) is declared as 1 dimensional, but later referred to as a 2D array- so I can't test your code.

[bluatigro]

update :
some last error's removed

i do not know how big e has to get before you get good results


'' bluatigro 16 nov 2018
'' ann bmp
'' based on :
''http://code.activestate.com/recipes/578148-simple-back-propagation-neural-network-in-python-s/
global ni , nh , no , size , hw , hdc
size = 63
ni = size * size
nh = size * sqr( size )
no = 1
dim ai( in( size , size ) ) , ah( nh )
dim ao( no ) , wish( no )
dim wi( ni , nh )
dim wo( nh , no )
dim ci( ni , nh )
dim co( nh , no )
dim od( nh ) , hd( nh )
WindowWidth = 800
WindowHeight = 600
open "ann bmp" for graphics as #m
 #m "trapclose [quit]"
 #m "font 50 bold"
 call init
 hw = hwnd( #m )
 calldll #user32, "GetDC" _
 , hw as ulong _   'handle of window or graphicbox client area
 , hdc as ulong    'returns handle of Device Context - 0=failure

 ''let NN live and learn
 for e = 0 to 100
   #m "fill black"
   #m "goto 0 " ; size + 200
   #m "down"
   #m "\live : " ; e
   #m "up"    
   #m "goto 0 " ; size + 300
   #m "down"
   #m "\error : " ; fout
   #m "up"        
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   ''fil target
   wish( 0 ) = dice
   call calc
   fout = backprop( .5 , .5 )
 next e
 notice "Lerning ready ."
 fout = 0
 for e = 0 to 100
   scan
   dice = drawsquare()
   ''fill input cel's
   for x = 0 to size
     for y = 0 to size
       if getpixel( hdc , x , y ) then
         ai( in( x , y ) ) = 1
       else
         ai( in( x , y ) ) = -1
       end if
     next y
   next x
   call calc
   if dice > 0 then
     if ao( 0 ) <= 0 then
       fout = fout + 1
     end if
   else
     if ao( 0 ) > 0 then
       fout = fout + 1
     end if
   end if
 next e
 #m "goto 0 " ; size + 50
 #m "down"
 #m "\error : " ; fout
 #m "up"
 notice "[ game over ]"
wait
[quit]
 calldll #user32 , "ReleaseDC" _
 , hw as ulong _    'window handle
 , hdc as ulong _  'device context
 , ret as long     'nonzero=success
 close #m
end
function getpixel( hDC , x , y )
 calldll #gdi32 , "GetPixel" _
 , hDC as ulong _
 , x as long _
 , y as long _
 , getpixel as ulong
end function
function in( x , y )
 in = x + y * size
end function
function drawsquare()
 r = irange( 10 , 30 )
 x = irange( r + 5 , 63 - r - 5 )
 y = irange( r + 5 , 63 - r - 5 )
 if irange( 0 , 1 ) then
   uit = 1
   #m "goto " ; x - r ; " " ; y - r
   #m "down"
   #m "boxfilled " ; x + r ; " " ; y + r
   #m "up"
 else
   uit = -1
   #m "goto " ; x ; " " ; y
   #m "down"
   #m "circlefilled " ; r
   #m "up"
 end if
 drawsquare = uit
end function
function irange( l , h )
 irange = int( range( l , h + 1 ) )
end function
function range( l , h )
 range = rnd(0) * ( h - l ) + l
end function
sub init
''init neural net
 for i = 0 to ni
   ai( i ) = 1
 next i
 for i = 0 to no - 1
   ao( i ) = 1
 next i
 for i = 0 to ni
   for h = 0 to nh - 1
     wi( i , h ) = range( -1 , 1 )
   next h
 next i
 for h = 0 to nh - 1
   for o = 0 to no - 1
     wo( h , o ) = range( -1 , 1 )
   next o
 next h
end sub
sub calc
''forwart pass of neural net
 for h = 0 to nh
   sum = 0
   for i = 0 to ni
     sum = sum + ai( i ) * wi( i , h )
   next i
   ah( h ) = signoid( sum / ni )
 next h
 for o = 0 to no
   sum = 0
   for h = 0 to nh
     sum = sum + ah( h ) * wo( h , o )
   next h
   ao( o ) = signoid( sum / nh )
 next o
end sub
function tanh( x )
 tanh = ( 1 - exp( -2 * x ) ) _
 / ( 1 + exp( -2 *x ) )
end function
function signoid( x )
 signoid = tanh( x )
end function
function dsignoid( x )
 dsignoid = 1 - x ^ 2
end function
function backprop( n , m )
'' http://www.youtube.com/watch?v=aVId8KMsdUU&feature=BFa&list=LLldMCkmXl4j9_v0HeKdNcRA
'' calc output deltas
'' we want to find the instantaneous rate of change of ( error with respect to weight from node j to node k)
'' output_delta is defined as an attribute of each ouput node. It is not the final rate we need.
'' To get the final rate we must multiply the delta by the activation of the hidden layer node in question.
'' This multiplication is done according to the chain rule as we are taking the derivative of the activation function
'' of the ouput node.
'' dE/dw[j][k] = (t[k] - ao[k]) * s'( SUM( w[j][k]*ah[j] ) ) * ah[j]
 for k = 0 to no - 1
   fout = wish( k ) - ao( k )
   od( k ) = fout * dsignoid( ao( k ) )
 next k
'' update output weights
 for j = 0 to nh
   for k = 0 to no
'' output_deltas[k] * self.ah[j]
'' is the full derivative of
'' dError/dweight[j][k]
     c = od( k ) * ah( j )
     wo( j , k ) = wo( j , k ) + n * c + m * co( j , k )
     co( j , k ) = c
   next k
 next j
'' calc hidden deltas input layer
 for j = 0 to nh
   fout = 0
   for k = 0 to no
     fout = fout + od( k ) * wo( j , k )
   next k
   hd( j ) = fout * dsignoid( ah( j ) )
 next j
'' update input weights
 for i = 0 to ni
   for j = 0 to nh
     c = hd( j ) * ai( i )
     wi( i , j ) = wi( i , j ) _
     + n * c + m * ci( i , j )
     ci( i , j ) = c
   next j
 next i
 fout = 0
 for k = 0 to no
   fout = fout _
   + ( wish( k ) - ao( k ) ) ^ 2
 next k
 backprop = fout / 2
end function