static int nn_(Min_updateOutput)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
int dimension = luaT_getfieldcheckint(L, 1, "dimension")-1;
THTensor *indices = luaT_getfieldcheckudata(L, 1, "indices", torch_(Tensor_id));
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THLongStorage *dim;
long i;
luaL_argcheck(L, dimension >= 0 && dimension < input->nDimension, 2, "dimension out of range");
dim = THLongStorage_newWithSize(input->nDimension);
for(i = 0; i < input->nDimension; i++)
dim->data[i] = input->size[i];
dim->data[dimension] = 1;
THTensor_(resize)(output, dim, NULL);
THTensor_(resize)(indices, dim, NULL);
THLongStorage_free(dim);
TH_TENSOR_DIM_APPLY3(real, output, real, input, real, indices, dimension,
long theIndex = 0;
real theMin = input_data[0];
for(i = 1; i < input_size; i++)
{
if(input_data[i*input_stride] < theMin)
{
theIndex = i;
theMin = input_data[i*input_stride];
}
}
*indices_data = theIndex+1;
*output_data = theMin;)
static int torch_(Tensor_copy)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
void *src;
if( (src = luaT_toudata(L, 2, torch_(Tensor_id))) )
THTensor_(copy)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_ByteTensor_id)) )
THTensor_(copyByte)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_CharTensor_id)) )
THTensor_(copyChar)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_ShortTensor_id)) )
THTensor_(copyShort)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_IntTensor_id)) )
THTensor_(copyInt)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_LongTensor_id)) )
THTensor_(copyLong)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_FloatTensor_id)) )
THTensor_(copyFloat)(tensor, src);
else if( (src = luaT_toudata(L, 2, torch_DoubleTensor_id)) )
THTensor_(copyDouble)(tensor, src);
else
luaL_typerror(L, 2, "torch.*Tensor");
lua_settop(L, 1);
return 1;
}
static int torch_(Tensor___index__)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
if(lua_isnumber(L, 2))
{
long index = luaL_checklong(L,2)-1;
luaL_argcheck(L, THTensor_(nDimension)(tensor) > 0, 1, "empty tensor");
if(THTensor_(nDimension)(tensor) == 1)
lua_pushnumber(L, THTensor_(get1d)(tensor, index));
else
{
THTensor *tensor_ = THTensor_(newWithTensorSelect)(tensor, index);
luaT_pushudata(L, tensor_, torch_(Tensor_id));
}
lua_pushboolean(L, 1);
return 2;
}
else
{
lua_pushboolean(L, 0);
return 1;
}
}
static int nn_(SparseLinear_updateOutput)(lua_State *L)
{
long i;
THTensor * input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor * weight = luaT_getfieldcheckudata(L, 1, "weight", torch_(Tensor_id));
THTensor * bias = luaT_getfieldcheckudata(L, 1, "bias", torch_(Tensor_id));
THTensor * output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
long dim = weight->size[0]; /* number of weights.. */
THTensor_(copy)(output, bias);
for(i = 0; i < input->size[1]; i++)
{
long offset = (long)(THTensor_(get2d)(input, 0, i))-1;
if(offset >= 0 && offset < dim) /* make sure indices are in bounds.. */
{
real val = THTensor_(get2d)(input, 1, i);
THBlas_(axpy)(output->size[0],
val,
THTensor_(data)(weight)+offset*weight->stride[0],
weight->stride[1],
THTensor_(data)(output),
output->stride[0]);
}
else
luaL_error(L, "index out of bound");
}
return 1;
}
static int torch_(Tensor_storage)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
THStorage_(retain)(THTensor_(storage)(tensor));
luaT_pushudata(L, THTensor_(storage)(tensor), torch_(Storage_id));
return 1;
}
/* Resize */
static int torch_(Tensor_resizeAs)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
THTensor *src = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor_(resizeAs)(tensor, src);
lua_settop(L, 1);
return 1;
}
static int torch_(Tensor_select)(lua_State *L)
{
THTensor *src = luaT_checkudata(L, 1, torch_(Tensor_id));
long sliceIndex = luaL_checklong(L, 2)-1;
THTensor *tensor = THTensor_(newWithTensorSelect)(src, sliceIndex);
luaT_pushudata(L, tensor, torch_(Tensor_id));
return 1;
}
static int nn_(Square_updateOutput)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor_(resizeAs)(output, input);
TH_TENSOR_APPLY2(real, output, real, input, \
*output_data = *input_data * *input_data;);
static int torch_(Tensor_narrow)(lua_State *L)
{
THTensor *src = luaT_checkudata(L, 1, torch_(Tensor_id));
long firstIndex = luaL_checklong(L, 2)-1;
long size = luaL_checklong(L, 3);
THTensor *tensor = THTensor_(newWithTensorNarrow)(src, firstIndex, size);
luaT_pushudata(L, tensor, torch_(Tensor_id));
return 1;
}
static int nn_(Sqrt_updateOutput)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
real bias = luaT_getfieldchecknumber(L,1,"eps");
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor_(resizeAs)(output, input);
TH_TENSOR_APPLY2(real, output, real, input, \
*output_data = sqrt(*input_data + bias););
static int nnOmp_(SpatialMaxPooling_updateGradInputOmp)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor *gradOutput = luaT_checkudata(L, 3, torch_(Tensor_id));
int kW = luaT_getfieldcheckint(L, 1, "kW");
int kH = luaT_getfieldcheckint(L, 1, "kH");
int dW = luaT_getfieldcheckint(L, 1, "dW");
int dH = luaT_getfieldcheckint(L, 1, "dH");
setompnthread(L,1,"nThread");
THTensor *indices = luaT_getfieldcheckudata(L, 1, "indices", torch_(Tensor_id));
THTensor *gradInput = luaT_getfieldcheckudata(L, 1, "gradInput", torch_(Tensor_id));
THTensor *gradOutputPlane, *gradInputPlane, *unfoldedGradInputPlane, *gradLocalInput;
int k,i,j;
THTensor_(resizeAs)(gradInput, input);
THTensor_(zero)(gradInput);
gradInputPlane = THTensor_(new)();
gradOutputPlane = THTensor_(new)();
gradLocalInput = THTensor_(new)();
unfoldedGradInputPlane = THTensor_(new)();
for (k = 0; k < input->size[0]; k++)
{
/* get input and output plane */
THTensor_(select)(gradOutputPlane, gradOutput, 0, k);
THTensor_(select)(gradInputPlane, gradInput, 0, k);
/* Unfold input to get each local window */
THTensor_(unfold)(unfoldedGradInputPlane, gradInputPlane, 0, kH, dH);
THTensor_(unfold)(unfoldedGradInputPlane, NULL, 1, kW, dW);
/* Calculate max points */
for(i = 0; i < gradOutputPlane->size[0]; i++) {
for(j = 0; j < gradOutputPlane->size[1]; j++) {
THTensor_(select)(gradLocalInput, unfoldedGradInputPlane,0,i);
THTensor_(select)(gradLocalInput, NULL, 0,j);
long maxi = THTensor_(get4d)(indices,0,k,i,j)-1;
long maxj = THTensor_(get4d)(indices,1,k,i,j)-1;
double gi = THTensor_(get2d)(gradLocalInput,maxi,maxj)+THTensor_(get2d)(gradOutputPlane,i,j);
THTensor_(set2d)(gradLocalInput,maxi,maxj,gi);
}
}
}
/* Cleanup */
THTensor_(free)(gradInputPlane);
THTensor_(free)(gradOutputPlane);
THTensor_(free)(unfoldedGradInputPlane);
THTensor_(free)(gradLocalInput);
return 1;
}
static int nn_(HardShrink_updateOutput)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor_(resizeAs)(output, input);
TH_TENSOR_APPLY2(real, output, real, input, \
if ((*input_data) > 0.5) *output_data = *input_data - 0.5; \
else if ((*input_data) < 0.5) *output_data = *input_data + 0.5; \
else *output_data = 0;);
static int nnOmp_(Tanh_updateOutputOmp)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
setompnthread(L,1,"nThread");
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor_(resizeAs)(output, input);
if (input->nDimension == 1 || !THTensor_(isContiguous)(input) || !THTensor_(isContiguous)(output))
{
TH_TENSOR_APPLY2(real, output, real, input, \
*output_data = tanh(*input_data););
static int torch_(Tensor_resize)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
THLongStorage *size;
torch_(Tensor_c_readSize)(L, 2, &size);
THTensor_(resize)(tensor, size);
THLongStorage_free(size);
lua_settop(L, 1);
return 1;
}
static void torch_(Tensor_c_readTensorStorageSize)(lua_State *L, int index, int allowNone, int allowTensor, int allowStorage,
THStorage **storage_, long *storageOffset_, THLongStorage **size_)
{
static char errMsg[64];
THTensor *src = NULL;
THStorage *storage = NULL;
int arg1Type = lua_type(L, index);
if( allowNone && (arg1Type == LUA_TNONE) )
{
*storage_ = NULL;
*storageOffset_ = 0;
*size_ = THLongStorage_new();
return;
}
else if( allowTensor && (arg1Type == LUA_TUSERDATA) && (src = luaT_toudata(L, index, torch_(Tensor_id))) )
{
*storage_ = THTensor_(storage)(src);
*storageOffset_ = THTensor_(storageOffset)(src);
*size_ = THTensor_(newSizeOf)(src);
return;
}
else if( allowStorage && (arg1Type == LUA_TUSERDATA) && (storage = luaT_toudata(L, index, torch_(Storage_id))) )
{
*storage_ = storage;
if(lua_isnone(L, index+1))
{
*storageOffset_ = 0;
*size_ = THLongStorage_newWithSize(1);
THLongStorage_set(*size_, 1, THStorage_(size)(storage));
}
else
{
*storageOffset_ = luaL_checklong(L, index+1)-1;
torch_(Tensor_c_readSize)(L, index+2, size_);
}
return;
}
else if( (arg1Type == LUA_TNUMBER) || (luaT_toudata(L, index, torch_LongStorage_id)) )
{
*storage_ = NULL;
*storageOffset_ = 0;
torch_(Tensor_c_readSize)(L, index, size_);
return;
}
sprintf(errMsg, "expecting number%s%s", (allowTensor ? " or Tensor" : ""), (allowStorage ? " or Storage" : ""));
luaL_argcheck(L, 0, index, errMsg);
}
static int torch_(Tensor_new)(lua_State *L)
{
THTensor *tensor;
THStorage *storage = NULL;
long storageOffset = 0;
THLongStorage *size = NULL;
torch_(Tensor_c_readTensorStorageSize)(L, 1, 1, 1, 1,
&storage, &storageOffset, &size);
tensor = THTensor_(newWithStorage)(storage, storageOffset, size);
THLongStorage_free(size);
luaT_pushudata(L, tensor, torch_(Tensor_id));
return 1;
}
/*******************
grab the rgb frame
*******************/
static int libkinect_(grab_rgb) (lua_State *L) {
// Get Tensor's Info
THTensor * tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
THTensor *contigTensor = THTensor_(newContiguous)(tensor);
// Get device ID
int index = 0;
if (lua_isnumber(L, 2)) index = lua_tonumber(L, 2);
THArgCheck(tensor->nDimension == 3 , 1, "RBG buffer: 3x480x640 Tensor expected");
THArgCheck(tensor->size[0] == 3 , 1, "RBG buffer: 3x480x640 Tensor expected");
THArgCheck(tensor->size[1] == 480 , 1, "RBG buffer: 3x480x640 Tensor expected");
THArgCheck(tensor->size[2] == 640 , 1, "RBG buffer: 3x480x640 Tensor expected");
unsigned int timestamp;
unsigned char *data = 0;
if (freenect_sync_get_video((void**)&data, ×tamp, index, FREENECT_VIDEO_RGB))
luaL_error(L, "<libkinect.grabRGB> Error Kinect not connected?");
int z;
for (z=0;z<3;z++){
unsigned char *sourcep = data+z;
THTensor *tslice = THTensor_(newSelect)(contigTensor,0,z);
// copy
TH_TENSOR_APPLY(real, tslice,
*tslice_data = ((real)(*sourcep)) / 255;
sourcep = sourcep + 3;
);
THTensor_(free)(tslice);
}
static int etherflow_(Api_send_tensor_lua)(lua_State *L) {
/* get the arguments */
THTensor *tensor = luaT_toudata(L, 1, torch_(Tensor_id));
int size = THTensor_(nElement)(tensor);
real *data = THTensor_(data)(tensor);
etherflow_send_(Tensor_C)(data, size);
return 0;
}
static int nn_(SpatialUpSampling_backward)(lua_State *L)
{
// get all params
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor *gradOutput = luaT_checkudata(L, 3, torch_(Tensor_id));
THTensor *gradInput = luaT_getfieldcheckudata(L, 1, "gradInput", torch_(Tensor_id));
int dW = luaT_getfieldcheckint(L, 1, "dW");
int dH = luaT_getfieldcheckint(L, 1, "dH");
// dims
int iwidth = input->size[2];
int iheight = input->size[1];
int ichannels = input->size[0];
int owidth = gradOutput->size[2];
int oheight = gradOutput->size[1];
int ochannels = gradOutput->size[0];
// resize gradInput
THTensor_(zero)(gradInput);
// get raw pointers
real *gradInput_data = THTensor_(data)(gradInput);
real *gradOutput_data = THTensor_(data)(gradOutput);
// compute gradients for each plane
int k;
for (k=0; k<ochannels; k++) {
// get planes
real *gradInput_p = gradInput_data + k*iwidth*iheight;
real *gradOutput_p = gradOutput_data + k*owidth*oheight;
// for each plane, resample
int x,y;
for (y=0; y<oheight; y++) {
for (x=0; x<owidth; x++) {
// input positions (floored)
int ix = x/dW;
int iy = y/dH;
// accumulate gradient
gradInput_p[iy*iwidth + ix] += gradOutput_p[y*owidth + x];
}
}
}
return 1;
}
static int nn_(TemporalMaxPooling_updateGradInput)(lua_State *L)
{
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor *gradOutput = luaT_checkudata(L, 3, torch_(Tensor_id));
int dW = luaT_getfieldcheckint(L, 1, "dW");
THTensor *indices = luaT_getfieldcheckudata(L, 1, "indices", torch_(Tensor_id));
THTensor *gradInput = luaT_getfieldcheckudata(L, 1, "gradInput", torch_(Tensor_id));
// get contiguous gradOutput
gradOutput = THTensor_(newContiguous)(gradOutput);
// resize and zero
THTensor_(resizeAs)(gradInput, input);
THTensor_(zero)(gradInput);
// sizes
int noframe = gradOutput->size[0];
long framesize = gradOutput->size[1];
// get raw pointers
real *gradInput_data = THTensor_(data)(gradInput);
real *gradOutput_data = THTensor_(data)(gradOutput);
real *indices_data = THTensor_(data)(indices);
long t, y;
for(t = 0; t < noframe; t++)
{
real *gip = gradInput_data + t*framesize*dW;
real *gop = gradOutput_data + t*framesize;
real *xp = indices_data + t*framesize;
#pragma omp parallel for private(y)
for(y = 0; y < framesize; y++)
{
// compute local max:
long maxindex = (long)xp[y];
gip[maxindex*framesize+y] += gop[y];
}
}
// cleanup
THTensor_(free)(gradOutput);
return 1;
}
static int nn_(SpatialUpSampling_forward)(lua_State *L)
{
// get all params
THTensor *input = luaT_checkudata(L, 2, torch_(Tensor_id));
int dW = luaT_getfieldcheckint(L, 1, "dW");
int dH = luaT_getfieldcheckint(L, 1, "dH");
THTensor *output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
// dims
int iwidth = input->size[2];
int iheight = input->size[1];
int ochannels = input->size[0];
int owidth = iwidth * dW;
int oheight = iheight * dH;
// get raw pointers
real *input_data = THTensor_(data)(input);
real *output_data = THTensor_(data)(output);
// resample each plane
int k;
for (k=0; k<ochannels; k++) {
// get planes
real *input_p = input_data + k*iwidth*iheight;
real *output_p = output_data + k*owidth*oheight;
// for each plane, resample
int x,y;
for (y=0; y<oheight; y++) {
for (x=0; x<owidth; x++) {
// input positions (floored)
int ix = x/dW;
int iy = y/dH;
// set output
output_p[y*owidth + x] = input_p[iy*iwidth + ix];
}
}
}
return 1;
}
int nn_(SparseLinear_updateParameters)(lua_State *L)
{
long i;
real learningRate = luaL_checknumber(L, 2);
THTensor * weight = luaT_getfieldcheckudata(L, 1, "weight", torch_(Tensor_id));
THTensor * output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor * bias = luaT_getfieldcheckudata(L, 1, "bias", torch_(Tensor_id));
THTensor * gradBias = luaT_getfieldcheckudata(L, 1, "gradBias", torch_(Tensor_id));
THTensor * gradWeight = luaT_getfieldcheckudata(L, 1, "gradWeight", torch_(Tensor_id));
THTensor * lastInput = luaT_getfieldcheckudata(L, 1, "lastInput", torch_(Tensor_id));
real weightDecay = luaT_getfieldchecknumber(L, 1, "weightDecay");
long dim = weight->size[0]; /* number of weights.. */
THTensor_(cadd)(bias, bias, -learningRate, gradBias);
for(i = 0; i < lastInput->size[1]; i++)
{
long offset = (long)(THTensor_(get2d)(lastInput, 0, i))-1;
if(offset >= 0 && offset < dim) /* make sure indices are in bounds.. */
{
THBlas_(axpy)(bias->size[0],
-learningRate,
THTensor_(data)(gradWeight)+offset*gradWeight->stride[0],
gradWeight->stride[1],
THTensor_(data)(weight)+offset*weight->stride[0],
weight->stride[1]);
}
else
luaL_error(L, "index out of bound");
}
return 0;
}
static int torch_(Tensor_size)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
if(lua_isnumber(L,2))
{
int dim = luaL_checkint(L, 2)-1;
luaL_argcheck(L, dim >= 0 && dim < THTensor_(nDimension)(tensor), 2, "out of range");
lua_pushnumber(L, THTensor_(size)(tensor, dim));
}
else
luaT_pushudata(L, THTensor_(newSizeOf)(tensor), torch_LongStorage_id);
return 1;
}
static int torch_(Tensor___newindex__)(lua_State *L)
{
if(lua_isnumber(L, 2))
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
long index = luaL_checklong(L,2)-1;
real value = (real)luaL_checknumber(L,3);
luaL_argcheck(L, THTensor_(nDimension)(tensor) == 1, 1, "must be a one dimensional tensor");
THTensor_(set1d)(tensor, index, value);
lua_pushboolean(L, 1);
}
else
lua_pushboolean(L, 0);
return 1;
}
static int nn_(SparseLinear_accGradParameters)(lua_State *L)
{
long i;
THTensor * input = luaT_checkudata(L, 2, torch_(Tensor_id));
THTensor * gradOutput = luaT_checkudata(L, 3, torch_(Tensor_id));
real scale = luaL_optnumber(L, 4, 1);
THTensor * weight = luaT_getfieldcheckudata(L, 1, "weight", torch_(Tensor_id));
THTensor * output = luaT_getfieldcheckudata(L, 1, "output", torch_(Tensor_id));
THTensor * gradBias = luaT_getfieldcheckudata(L, 1, "gradBias", torch_(Tensor_id));
THTensor * gradWeight = luaT_getfieldcheckudata(L, 1, "gradWeight", torch_(Tensor_id));
THTensor * lastInput = luaT_getfieldcheckudata(L, 1, "lastInput", torch_(Tensor_id));
real weightDecay = luaT_getfieldchecknumber(L, 1, "weightDecay");
long dim = gradWeight->size[0]; /* number of weights.. */
for(i = 0; i < input->size[1]; i++)
{
long offset = (long)(THTensor_(get2d)(input, 0, i))-1;
if(offset >= 0 && offset < dim) /* make sure indices are in bounds.. */
{
real val = scale*THTensor_(get2d)(input, 1, i);
THBlas_(scal)(gradOutput->size[0],
0,
THTensor_(data)(gradWeight)+offset*gradWeight->stride[0],
gradWeight->stride[1]); /* zero */
THBlas_(axpy)(gradOutput->size[0],
val,
THTensor_(data)(gradOutput),
gradOutput->stride[0],
THTensor_(data)(gradWeight)+offset*gradWeight->stride[0],
gradWeight->stride[1]);
}
else
luaL_error(L, "index out of bound");
}
THTensor_(cadd)(gradBias, gradBias, 1, gradOutput);
if(weightDecay != 0)
THTensor_(cadd)(gradWeight, gradWeight, weightDecay, weight);
THTensor_(resizeAs)(lastInput, input);
THTensor_(copy)(lastInput, input);
return 0;
}
static int torch_(Tensor_apply)(lua_State *L)
{
THTensor *tensor = luaT_checkudata(L, 1, torch_(Tensor_id));
luaL_checktype(L, 2, LUA_TFUNCTION);
lua_settop(L, 2);
TH_TENSOR_APPLY(tensor,
lua_pushvalue(L, 2);
lua_pushnumber(L, tensor_data[i]);
lua_call(L, 1, 1);
if(lua_isnumber(L, 3))
{
tensor_data[i] = (real)lua_tonumber(L, 3);
lua_pop(L, 1);
}
else if(lua_isnil(L, 3))
lua_pop(L, 1);
else
luaL_error(L, "given function should return a number or nil"););
void etherflow_(Api_init)(lua_State *L)
{
luaT_pushmetaclass(L, torch_(Tensor_id));
luaT_registeratname(L, etherflow_(Api__), "etherflow");
}
// Stitch takes args.
// pano - a torch tensor in RGB with dims (3 x height x width)
//
// offset_map - a torch tensor same h and w as pano storing offsets and
// and image indices. The two feaure dimentions are:
// -- image number (starting at 1) and
// -- bit offset in image tensor
// nimages - is the number of images use to make the panorama
// image1, ... imagen - are the image in a torch tensor
static int Lstitch_(stitch)(lua_State *L) {
int nargs = lua_gettop(L);
THTensor *pano =
(THTensor *)luaT_checkudata(L, 1, torch_(Tensor_id));
THLongTensor *offset_map =
(THLongTensor *)luaT_checkudata(L, 2, luaT_checktypename2id(L, "torch.LongTensor"));
THTensor *images[MAXIMAGES];
int i = 0;
long npixels = offset_map->size[1]*offset_map->size[2];
real *pano_pt = THTensor_(data)(pano);
long *offset_pt = THLongTensor_data(offset_map);
real *images_pt[MAXIMAGES];
long images_npixels[MAXIMAGES];
long images_Goff[MAXIMAGES];
long images_Boff[MAXIMAGES];
real * panoR = pano_pt;
real * panoG = pano_pt + pano->stride[0];
real * panoB = pano_pt + (2*pano->stride[0]);
real * curImg_pt = NULL;
long unsigned int XYoffset = 0;
long * offImg = offset_pt;
long * offIndexXY = offset_pt + offset_map->stride[0];
int nimages = 0;
long cImgOff = 0;
/* finish processing input image tensors */
/* either you can pass a table */
/* or a number and variable length of args */
if (nargs == 3){
if (lua_istable(L,3)){
nimages = lua_objlen (L, 3);
/* table is in the stack at index 3 */
lua_pushnil(L); /* first key */
i = 0;
while (lua_next(L, 3) != 0) {
/* 'key' (at index -2) and 'value' (at index -1) */
images[i] =
(THTensor *)luaT_checkudata(L, -1, torch_(Tensor_id));
images_npixels[i] = images[i]->size[1]*images[i]->size[2];
images_Goff[i] = images[i]->stride[0];
images_Boff[i] = 2*images[i]->stride[0];
images_pt[i] = THTensor_(data)(images[i]);
/* removes 'value'; keeps 'key' for next iteration */
lua_pop(L, 1);
i = i+1;
}
} else {
lua_pushstring(L, "with 3 args last argument is a table");
lua_error(L);
}
} else {
nimages = lua_tonumber(L,3);
for(i=0;i<nimages;i++){
images[i] =
(THTensor *)luaT_checkudata(L, i+4, torch_(Tensor_id));
images_npixels[i] = images[i]->size[1]*images[i]->size[2];
images_Goff[i] = images[i]->stride[0];
images_Boff[i] = 2*images[i]->stride[0];
images_pt[i] = THTensor_(data)(images[i]);
}
}
for(i=0;i<npixels;i++){
cImgOff = (long unsigned int)*offImg - 1;
curImg_pt = images_pt[cImgOff];
if ((*offIndexXY > 0) &&
(*offIndexXY < images_npixels[cImgOff])){
XYoffset = (long unsigned int)*offIndexXY;
*panoR = curImg_pt[XYoffset];
*panoG = curImg_pt[XYoffset + images_Goff[cImgOff]] ;
*panoB = curImg_pt[XYoffset + images_Boff[cImgOff]];
}
panoR++;
panoG++;
panoB++;
offImg++;
offIndexXY++;
}
return 0;
}
void Lstitch_(Init)(lua_State *L)
{
luaT_pushmetaclass(L, torch_(Tensor_id));
luaT_registeratname(L, Lstitch_(Methods), "stitch");
}
static void nn_(SpatialUpSampling_init)(lua_State *L)
{
luaT_pushmetaclass(L, torch_(Tensor_id));
luaT_registeratname(L, nn_(SpatialUpSampling__), "nn");
lua_pop(L,1);
}
