Brutex
/
cavis
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Shugeo release fix1 (#61) 

* Added a pair of tests for failed ops.

* Fixed cpu helper for draw_bounding_boxes op.

* Refactored implementation of draw_bounding_boxes op to full conform with TF.

* Improved multithreading with draw_bounding_boxes op cuda helper.

* Eliminated log messages.

* Changed logging with draw_bounding_boxes op helper and tests.

* Resize_biliear with 3D input allowed.

* Refactored 3D input acception with resize_bilinear op.

* And another improvement.

* Refactored reshape of input/output for resize_bilinear.

* Improvements final.

* Finished with 3D replication for image.resize_bilinear/_nearest_neighbor.

* Added copyrights for TF code.

* Using new form of multithreading for cpu implementation.

* Fixed shape error.

* Added multithreaded with batches on crop_and_resize functor.

* Refactored multithreading with crop_and_resize and draw_bounding_boxes.
master
shugeo 2019-11-20 13:37:48 +02:00 committed by GitHub
parent 59e955cedc
commit 13e5c0a280
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
9 changed files with 565 additions and 106 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
libnd4j/include/ops/declarable/generic/parity_ops/crop_and_resize.cpp
View File

@ -57,6 +57,7 @@ namespace nd4j {
DECLARE_SHAPE_FN(crop_and_resize) {
auto in = inputShape->at(0);
auto boxShape = inputShape->at(1);
Nd4jLong outputShape[4];
@ -68,7 +69,7 @@ namespace nd4j {
width = newImageSize->e<int>(0);
height = newImageSize->e<int>(1);
outputShape[0] = in[1];
outputShape[0] = boxShape[1];
outputShape[1] = width;
outputShape[2] = height;
outputShape[3] = in[4];

26
libnd4j/include/ops/declarable/generic/parity_ops/draw_bounding_boxes.cpp
View File

@ -29,9 +29,31 @@ namespace nd4j {
auto images = INPUT_VARIABLE(0);
auto boxes = INPUT_VARIABLE(1);
auto colors = INPUT_VARIABLE(2);
auto output = OUTPUT_VARIABLE(0);
auto colors = (NDArray*) nullptr;
if (block.width() > 2) // TF v.1.x ommits color set for boxes, and use color 1.0 for fill up
colors = INPUT_VARIABLE(2); // but v.2.y require color set
auto output = OUTPUT_VARIABLE(0);
REQUIRE_TRUE(images->dataType() == output->dataType(), 0, "draw_bounding_boxes: Input and Output types "
"should be equals, but %d and %d occured.",
(int)images->dataType(), (int)output->dataType());
REQUIRE_TRUE(images->rankOf() == 4, 0, "draw_bounding_boxes: Images input should be 4D tensor, but %i occured.",
images->rankOf());
REQUIRE_TRUE(boxes->rankOf() == 3, 0, "draw_bounding_boxes: Boxes should be 3D tensor, but %i occured.",
boxes->rankOf());
if (colors) {
REQUIRE_TRUE(colors->rankOf() == 2, 0, "draw_bounding_boxes: Color set should be 2D matrix, but %i occured.",
colors->rankOf());
REQUIRE_TRUE(colors->sizeAt(1) >= images->sizeAt(3), 0, "draw_bounding_boxes: Color set last dim "
"should be not less than images depth, but "
"%lld and %lld occured.",
colors->sizeAt(1), images->sizeAt(3));
}
REQUIRE_TRUE(boxes->sizeAt(0) == images->sizeAt(0), 0, "draw_bounding_boxes: Batches for images and boxes "
"should be the same, but %lld and %lld occured.",
images->sizeAt(0), boxes->sizeAt(0));
helpers::drawBoundingBoxesFunctor(block.launchContext(), images, boxes, colors, output);
return ND4J_STATUS_OK;
}

35
libnd4j/include/ops/declarable/generic/parity_ops/resize_linear.cpp
View File

@ -33,6 +33,15 @@ namespace nd4j {
int width;
int height;
bool center = false; // - default value
auto inRank = image->rankOf();
REQUIRE_TRUE( inRank == 4 || inRank == 3, 0, "resize_bilinear: input image should be 4D "
"tensor, but input has rank %i",
image->rankOf());
REQUIRE_TRUE(inRank == output->rankOf(), 0, "resize_bilinear: Input and output ranks should be equals, but %i and %i occured.", inRank, output->rankOf());
auto source = inRank == 4?image->reshape(image->ordering(), {image->sizeAt(0), image->sizeAt(1), image->sizeAt(2), image->sizeAt(3)}):image->reshape(image->ordering(), {1, image->sizeAt(0), image->sizeAt(1), image->sizeAt(2)});
auto target = inRank == 4?output->reshape(output->ordering(), {output->sizeAt(0), output->sizeAt(1), output->sizeAt(2), output->sizeAt(3)}):output->reshape(output->ordering(), {1, output->sizeAt(0), output->sizeAt(1), output->sizeAt(2)});
if (block.width() > 1) {
auto newImageSize = INPUT_VARIABLE(1);
REQUIRE_TRUE(newImageSize->lengthOf() == 2, 0, "resize_bilinear: Resize params is a pair of values, not %i.", newImageSize->lengthOf());
@ -51,7 +60,8 @@ namespace nd4j {
center = 0 != INT_ARG(2);
}
return helpers::resizeBilinearFunctor(block.launchContext(), image, width, height, center, output);
auto res = helpers::resizeBilinearFunctor(block.launchContext(), &source, width, height, center, &target);
return res;
}
DECLARE_SHAPE_FN(resize_bilinear) {
@ -59,6 +69,10 @@ namespace nd4j {
auto in = inputShape->at(0);
Nd4jLong* outputShape;
auto inRank = shape::rank(in);
REQUIRE_TRUE(inRank == 4 || inRank == 3, 0, "resize_bilinear: input image should be 4D "
"tensor, but input has rank %i",
inRank);
int width;
int height;
@ -75,12 +89,19 @@ namespace nd4j {
height = INT_ARG(1);
}
ALLOCATE(outputShape, block.getWorkspace(), shape::shapeInfoLength(4), Nd4jLong);
outputShape[0] = 4;
outputShape[1] = in[1];
outputShape[2] = width;
outputShape[3] = height;
outputShape[4] = in[4];
ALLOCATE(outputShape, block.getWorkspace(), shape::shapeInfoLength(inRank), Nd4jLong);
outputShape[0] = inRank;
if (inRank == 4) {
outputShape[1] = in[1];
outputShape[2] = width;
outputShape[3] = height;
outputShape[4] = in[4];
}
else { // input shape is 3D, so result also should be 3D
outputShape[1] = width;
outputShape[2] = height;
outputShape[3] = in[3];
}
ShapeUtils::updateStridesAndType(outputShape, in, shape::order(in));
shapeList->push_back(CONSTANT(outputShape));

34
libnd4j/include/ops/declarable/generic/parity_ops/resize_neighbor.cpp
View File

@ -51,16 +51,25 @@ namespace nd4j {
if (block.numI() == 3)
center = 0 != INT_ARG(2);
}
auto inRank = image->rankOf();
REQUIRE_TRUE(inRank == 4 || inRank == 3, 0, "resize_nearest_neighbor: Input should be 4D tensor, but rank %i occured");
REQUIRE_TRUE(inRank == output->rankOf(), 0, "resize_nearest_neighbor: Input and output ranks should be equals, but %i and %i occured.", inRank, output->rankOf());
auto source = inRank == 4?image->reshape(image->ordering(), {image->sizeAt(0), image->sizeAt(1), image->sizeAt(2), image->sizeAt(3)}):image->reshape(image->ordering(), {1, image->sizeAt(0), image->sizeAt(1), image->sizeAt(2)});
auto target = inRank == 4?output->reshape(output->ordering(), {output->sizeAt(0), output->sizeAt(1), output->sizeAt(2), output->sizeAt(3)}):output->reshape(output->ordering(), {1, output->sizeAt(0), output->sizeAt(1), output->sizeAt(2)});
return helpers::resizeNeighborFunctor(block.launchContext(), image, width, height, center, output);
return helpers::resizeNeighborFunctor(block.launchContext(), &source, width, height, center, &target);
}
DECLARE_SHAPE_FN(resize_nearest_neighbor) {
auto shapeList = SHAPELIST();
auto in = inputShape->at(0);
auto inRank = shape::rank(in);
Nd4jLong* outputShape;
REQUIRE_TRUE(inRank == 4 || inRank == 3, 0, "resize_bilinear: input image should be 4D "
"tensor, but input has rank %i",
inRank);
int width;
int height;
if (block.width() > 1) {
@ -75,13 +84,20 @@ namespace nd4j {
width = INT_ARG(0);
height = INT_ARG(1);
}
ALLOCATE(outputShape, block.getWorkspace(), shape::shapeInfoLength(4), Nd4jLong);
outputShape[0] = 4;
outputShape[1] = in[1];
outputShape[2] = width;
outputShape[3] = height;
outputShape[4] = in[4];
ALLOCATE(outputShape, block.getWorkspace(), shape::shapeInfoLength(inRank), Nd4jLong);
outputShape[0] = inRank;
if (inRank == 4) {
outputShape[1] = in[1];
outputShape[2] = width;
outputShape[3] = height;
outputShape[4] = in[4];
}
else { // input shape is 3D, so result also should be 3D
outputShape[1] = width;
outputShape[2] = height;
outputShape[3] = in[3];
}
ShapeUtils::updateStridesAndType(outputShape, in, shape::order(in));
shapeList->push_back(CONSTANT(outputShape));

2
libnd4j/include/ops/declarable/helpers/cpu/fake_quantization.cpp
View File

@ -47,7 +47,7 @@ namespace helpers {
if (zeroPointFromMin > quantMaxF) {
return static_cast<uint16_t>(quantMax);
}
return nd4j::math::nd4j_round<T,uint16_t>(zeroPointFromMin);
return (uint16_t)nd4j::math::nd4j_round<T,int>(zeroPointFromMin);
}();
// compute nudged min and max with computed nudged zero point
*nudgedMin = (quantMinF - nudged_zero_point) * (*scale);

157
libnd4j/include/ops/declarable/helpers/cpu/image_draw_bounding_boxes.cpp
View File

@ -13,16 +13,52 @@
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
/* Copyright 2016 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
//
// @author sgazeos@gmail.com
//
#include <op_boilerplate.h>
#include <NDArray.h>
#include <execution/Threads.h>
namespace nd4j {
namespace ops {
namespace helpers {
typedef std::vector<std::vector<float>> ColorTable_t;
static ColorTable_t DefaultColorTable(int depth) {
std::vector<std::vector<float>> colorTable;
colorTable.emplace_back(std::vector<float>({1, 1, 0, 1})); // 0: yellow
colorTable.emplace_back(std::vector<float>({0, 0, 1, 1})); // 1: blue
colorTable.emplace_back(std::vector<float>({1, 0, 0, 1})); // 2: red
colorTable.emplace_back(std::vector<float>({0, 1, 0, 1})); // 3: lime
colorTable.emplace_back(std::vector<float>({0.5, 0, 0.5, 1})); // 4: purple
colorTable.emplace_back(std::vector<float>({0.5, 0.5, 0, 1})); // 5: olive
colorTable.emplace_back(std::vector<float>({0.5, 0, 0, 1})); // 6: maroon
colorTable.emplace_back(std::vector<float>({0, 0, 0.5, 1})); // 7: navy blue
colorTable.emplace_back(std::vector<float>({0, 1, 1, 1})); // 8: aqua
colorTable.emplace_back(std::vector<float>({1, 0, 1, 1})); // 9: fuchsia
if (depth == 1) {
for (Nd4jLong i = 0; i < colorTable.size(); i++) {
colorTable[i][0] = 1;
}
}
return colorTable;
}
void drawBoundingBoxesFunctor(nd4j::LaunchContext * context, NDArray* images, NDArray* boxes, NDArray* colors, NDArray* output) {
// images - batch of 3D images with BW (last dim = 1), RGB (last dim = 3) or RGBA (last dim = 4) channel set
@ -33,44 +69,107 @@ namespace helpers {
// colors - colors for each box given
// set up color for each box as frame
auto batchSize = images->sizeAt(0);
auto boxSize = boxes->sizeAt(0);
auto height = images->sizeAt(1);
auto width = images->sizeAt(2);
auto channels = images->sizeAt(3);
//auto imageList = images->allTensorsAlongDimension({1, 2, 3}); // split images by batch
// auto boxList = boxes->allTensorsAlongDimension({1, 2}); // split boxes by batch
auto colorSet = colors->allTensorsAlongDimension({1});
//auto colorSet = colors->allTensorsAlongDimension({0});
output->assign(images); // fill up all output with input images, then fill up boxes
PRAGMA_OMP_PARALLEL_FOR
for (auto b = 0; b < batchSize; ++b) { // loop by batch
// auto image = imageList->at(b);
for (auto c = 0; c < colorSet->size(); ++c) {
// box with shape
auto internalBox = (*boxes)(b, {0})(c, {0});//internalBoxes->at(c);
auto color = colorSet->at(c);
auto rowStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((height - 1) * internalBox.e<float>(0)));
auto rowEnd = nd4j::math::nd4j_min(Nd4jLong (height - 1), Nd4jLong ((height - 1) * internalBox.e<float>(2)));
auto colStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((width - 1) * internalBox.e<float>(1)));
auto colEnd = nd4j::math::nd4j_min(Nd4jLong(width - 1), Nd4jLong ((width - 1) * internalBox.e<float>(3)));
for (auto y = rowStart; y <= rowEnd; y++) {
for (auto e = 0; e < color->lengthOf(); ++e) {
output->p(b, y, colStart, e, color->e(e));
output->p(b, y, colEnd, e, color->e(e));
}
}
for (auto x = colStart + 1; x < colEnd; x++) {
for (auto e = 0; e < color->lengthOf(); ++e) {
output->p(b, rowStart, x, e, color->e(e));
output->p(b, rowEnd, x, e, color->e(e));
}
ColorTable_t colorTable;
if (colors) {
for (auto i = 0; i < colors->sizeAt(0); i++) {
std::vector<float> colorValue(4);
for (auto j = 0; j < 4; j++) {
colorValue[j] = j < colors->sizeAt(1) ? colors->e<float>(i, j) : 1.f;
}
colorTable.emplace_back(colorValue);
}
// delete internalBoxes;
}
delete colorSet;
// delete imageList;
// delete boxList;
if (colorTable.empty())
colorTable = DefaultColorTable(channels);
// PRAGMA_OMP_PARALLEL_FOR
auto func = PRAGMA_THREADS_FOR {
for (auto batch = 0; batch < batchSize; ++batch) { // loop by batch
// auto image = imageList->at(batch);
const Nd4jLong numBoxes = boxes->sizeAt(1);
for (auto boxIndex = 0; boxIndex < numBoxes; ++boxIndex) {
// box with shape
//auto internalBox = (*boxes)(batch, {0})(boxIndex, {0});//internalBoxes->at(c);
//auto color = colorSet->at(c);
//internalBox.printIndexedBuffer("Current Box");
auto colorIndex = boxIndex % colorTable.size();
auto rowStart = Nd4jLong((height - 1) * boxes->t<float>(batch, boxIndex, 0));
auto rowStartBound = nd4j::math::nd4j_max(Nd4jLong(0), rowStart);
auto rowEnd = Nd4jLong((height - 1) * boxes->t<float>(batch, boxIndex, 2));
auto rowEndBound = nd4j::math::nd4j_min(Nd4jLong(height - 1), rowEnd);
auto colStart = Nd4jLong((width - 1) * boxes->t<float>(batch, boxIndex, 1));
auto colStartBound = nd4j::math::nd4j_max(Nd4jLong(0), colStart);
auto colEnd = Nd4jLong((width - 1) * boxes->t<float>(batch, boxIndex, 3));
auto colEndBound = nd4j::math::nd4j_min(Nd4jLong(width - 1), colEnd);
if (rowStart > rowEnd || colStart > colEnd) {
nd4j_debug(
"helpers::drawBoundingBoxesFunctor: Bounding box (%lld, %lld, %lld, %lld) is inverted "
"and will not be drawn\n", rowStart, colStart, rowEnd, colEnd);
continue;
}
if (rowStart >= height || rowEnd < 0 || colStart >= width ||
colEnd < 0) {
nd4j_debug(
"helpers::drawBoundingBoxesFunctor: Bounding box (%lld, %lld, %lld, %lld) is completely "
"outside the image and not be drawn\n ", rowStart, colStart, rowEnd, colEnd);
continue;
}
// Draw upper line
if (rowStart >= 0) {
for (auto j = colStartBound; j <= colEndBound; ++j)
for (auto c = 0; c < channels; c++) {
output->p(batch, rowStart, j, c, colorTable[colorIndex][c]);
}
}
// Draw bottom line.
if (rowEnd < height) {
for (auto j = colStartBound; j <= colEndBound; ++j)
for (auto c = 0; c < channels; c++) {
output->p(batch, rowEnd, j, c, colorTable[colorIndex][c]);
}
}
// Draw left line.
if (colStart >= 0) {
for (auto i = rowStartBound; i <= rowEndBound; ++i)
for (auto c = 0; c < channels; c++) {
output->p(batch, i, colStart, c, colorTable[colorIndex][c]);
}
}
// Draw right line.
if (colEnd < width) {
for (auto i = rowStartBound; i <= rowEndBound; ++i)
for (auto c = 0; c < channels; c++) {
output->p(batch, i, colEnd, c, colorTable[colorIndex][c]);
}
}
// for (auto y = rowStart; y <= rowEnd; y++) {
// for (auto e = 0; e < channels; ++e) {
// output->p(batch, y, colStart, e, colorTable[colorIndex][e]);
// output->p(batch, y, colEnd, e, colorTable[colorIndex][e]);
// }
// }
// for (auto x = colStart + 1; x < colEnd; x++) {
// for (auto e = 0; e < channels; ++e) {
// output->p(batch, rowStart, x, e, colorTable[colorIndex][e]);
// output->p(batch, rowEnd, x, e, colorTable[colorIndex][e]);
// }
// }
}
// delete internalBoxes;
}
};
samediff::Threads::parallel_tad(func, 0, batchSize);
}
}

58
libnd4j/include/ops/declarable/helpers/cpu/image_resize.cpp
View File

@ -14,6 +14,20 @@
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
/* Copyright 2016 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
//
// @author sgazeos@gmail.com
@ -85,30 +99,32 @@ namespace helpers {
return top + (bottom - top) * yVal;
};
// FIXME: fix parallelism here
for (Nd4jLong b = 0; b < batchSize; ++b) {
for (Nd4jLong y = 0; y < outHeight; ++y) {
const T *ys_input_lower_ptr = input_b_ptr + ys[y].bottomIndex * inRowSize;
const T *ys_input_upper_ptr = input_b_ptr + ys[y].topIndex * inRowSize;
double yVal = ys[y].interpolarValue;
for (Nd4jLong x = 0; x < outWidth; ++x) {
auto xsBottom = xs_[x].bottomIndex;
auto xsTop = xs_[x].topIndex;
auto xVal = xs_[x].interpolarValue;
for (int c = 0; c < channels; ++c) {
double topLeft(ys_input_lower_ptr[xsBottom + c]);
double topRight(ys_input_lower_ptr[xsTop + c]);
double bottomLeft(ys_input_upper_ptr[xsBottom + c]);
double bottomRight(ys_input_upper_ptr[xsTop + c]);
output_y_ptr[x * channels + c] =
computeBilinear(topLeft, topRight, bottomLeft, bottomRight,
xVal, yVal);
auto func = PRAGMA_THREADS_FOR {
for (Nd4jLong b = 0; b < batchSize; ++b) {
for (Nd4jLong y = 0; y < outHeight; ++y) {
const T *ys_input_lower_ptr = input_b_ptr + ys[y].bottomIndex * inRowSize;
const T *ys_input_upper_ptr = input_b_ptr + ys[y].topIndex * inRowSize;
double yVal = ys[y].interpolarValue;
for (Nd4jLong x = 0; x < outWidth; ++x) {
auto xsBottom = xs_[x].bottomIndex;
auto xsTop = xs_[x].topIndex;
auto xVal = xs_[x].interpolarValue;
for (int c = 0; c < channels; ++c) {
double topLeft(ys_input_lower_ptr[xsBottom + c]);
double topRight(ys_input_lower_ptr[xsTop + c]);
double bottomLeft(ys_input_upper_ptr[xsBottom + c]);
double bottomRight(ys_input_upper_ptr[xsTop + c]);
output_y_ptr[x * channels + c] =
computeBilinear(topLeft, topRight, bottomLeft, bottomRight,
xVal, yVal);
}
}
output_y_ptr += outRowSize;
}
output_y_ptr += outRowSize;
input_b_ptr += inBatchNumValues;
}
input_b_ptr += inBatchNumValues;
}
};
samediff::Threads::parallel_tad(func, 0, batchSize);
}
template<typename T>

149
libnd4j/include/ops/declarable/helpers/cuda/image_draw_bounding_boxes.cu
View File

@ -24,39 +24,112 @@ namespace nd4j {
namespace ops {
namespace helpers {
template <typename T>
static __global__ void drawBoundingBoxesKernel(T const* images, Nd4jLong* imagesShape, T const* boxes,
Nd4jLong* boxesShape, T const* colors, Nd4jLong* colorsShape, T* output, Nd4jLong* outputShape,
Nd4jLong batchSize, Nd4jLong width, Nd4jLong height, Nd4jLong channels, Nd4jLong colorSetSize) {
typedef NDArray ColorTable_t;
static NDArray DefaultColorTable(int depth) {
//std::vector<std::vector<float>> colorTable;
const Nd4jLong kDefaultTableLength = 10;
const Nd4jLong kDefaultChannelLength = 4;
NDArray colorTable('c', {kDefaultTableLength, kDefaultChannelLength}, {
1,1,0,1, // yellow
0, 0, 1, 1, // 1: blue
1, 0, 0, 1, // 2: red
0, 1, 0, 1, // 3: lime
0.5, 0, 0.5, 1, // 4: purple
0.5, 0.5, 0, 1, // 5: olive
0.5, 0, 0, 1, // 6: maroon
0, 0, 0.5, 1, // 7: navy blue
0, 1, 1, 1, // 8: aqua
1, 0, 1, 1 // 9: fuchsia
}, DataType::FLOAT32);
for (auto b = blockIdx.x; b < (int)batchSize; b += gridDim.x) { // loop by batch
for (auto c = 0; c < colorSetSize; c++) {
if (depth == 1) {
colorTable.assign(1.f); // all to white when black and white colors
}
return colorTable;
}
template <typename T>
static __global__ void drawBoundingBoxesKernel(T const* images, Nd4jLong* imagesShape, float const* boxes,
Nd4jLong* boxesShape, float const* colorTable, Nd4jLong* colorTableShape, T* output, Nd4jLong* outputShape,
Nd4jLong batchSize, Nd4jLong width, Nd4jLong height, Nd4jLong channels, Nd4jLong boxSize, Nd4jLong colorTableLen) {
for (auto batch = blockIdx.x; batch < (int)batchSize; batch += gridDim.x) { // loop by batch
for (auto boxIndex = 0; boxIndex < boxSize; ++boxIndex) {
// box with shape
auto internalBox = &boxes[b * colorSetSize * 4 + c * 4];//(*boxes)(b, {0})(c, {0});//internalBoxes->at(c);
auto color = &colors[channels * c];//colorSet->at(c);
auto rowStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((height - 1) * internalBox[0]));
auto rowEnd = nd4j::math::nd4j_min(Nd4jLong (height - 1), Nd4jLong ((height - 1) * internalBox[2]));
auto colStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((width - 1) * internalBox[1]));
auto colEnd = nd4j::math::nd4j_min(Nd4jLong(width - 1), Nd4jLong ((width - 1) * internalBox[3]));
for (auto y = rowStart + threadIdx.x; y <= rowEnd; y += blockDim.x) {
for (auto e = 0; e < channels; ++e) {
Nd4jLong yMinPos[] = {b, y, colStart, e};
Nd4jLong yMaxPos[] = {b, y, colEnd, e};
auto zIndexYmin = shape::getOffset(outputShape, yMinPos);
auto zIndexYmax = shape::getOffset(outputShape, yMaxPos);
output[zIndexYmin] = color[e];
output[zIndexYmax] = color[e];
}
//auto internalBox = &boxes[b * colorSetSize * 4 + c * 4];//(*boxes)(b, {0})(c, {0});//internalBoxes->at(c);
auto colorIndex = boxIndex % colorTableLen;//colorSet->at(c);
// auto rowStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((height - 1) * internalBox[0]));
// auto rowEnd = nd4j::math::nd4j_min(Nd4jLong (height - 1), Nd4jLong ((height - 1) * internalBox[2]));
// auto colStart = nd4j::math::nd4j_max(Nd4jLong (0), Nd4jLong ((width - 1) * internalBox[1]));
// auto colEnd = nd4j::math::nd4j_min(Nd4jLong(width - 1), Nd4jLong ((width - 1) * internalBox[3]));
Nd4jLong indices0[] = {batch, boxIndex, 0};
Nd4jLong indices1[] = {batch, boxIndex, 1};
Nd4jLong indices2[] = {batch, boxIndex, 2};
Nd4jLong indices3[] = {batch, boxIndex, 3};
auto rowStart = Nd4jLong ((height - 1) * boxes[shape::getOffset(boxesShape, indices0, 0)]);
auto rowStartBound = nd4j::math::nd4j_max(Nd4jLong (0), rowStart);
auto rowEnd = Nd4jLong ((height - 1) * boxes[shape::getOffset(boxesShape, indices2, 0)]);
auto rowEndBound = nd4j::math::nd4j_min(Nd4jLong (height - 1), rowEnd);
auto colStart = Nd4jLong ((width - 1) * boxes[shape::getOffset(boxesShape, indices1, 0)]);
auto colStartBound = nd4j::math::nd4j_max(Nd4jLong (0), colStart);
auto colEnd = Nd4jLong ((width - 1) * boxes[shape::getOffset(boxesShape, indices3, 0)]);
auto colEndBound = nd4j::math::nd4j_min(Nd4jLong(width - 1), colEnd);
if (rowStart > rowEnd || colStart > colEnd) {
// printf("helpers::drawBoundingBoxesFunctor: Bounding box (%lld, %lld, %lld, %lld) is inverted "
// "and will not be drawn\n", rowStart, colStart, rowEnd, colEnd);
continue;
}
for (auto x = colStart + 1 + threadIdx.x; x < colEnd; x += blockDim.x) {
for (auto e = 0; e < channels; ++e) {
Nd4jLong xMinPos[] = {b, rowStart, x, e};
Nd4jLong xMaxPos[] = {b, rowEnd, x, e};
auto zIndexXmin = shape::getOffset(outputShape, xMinPos);
auto zIndexXmax = shape::getOffset(outputShape, xMaxPos);
output[zIndexXmin] = color[e];
output[zIndexXmax] = color[e];
}
if (rowStart >= height || rowEnd < 0 || colStart >= width ||
colEnd < 0) {
// printf("helpers::drawBoundingBoxesFunctor: Bounding box (%lld, %lld, %lld, %lld) is completely "
// "outside the image and not be drawn\n", rowStart, colStart, rowEnd, colEnd);
continue;
}
// Draw upper line
if (rowStart >= 0) {
for (auto j = colStartBound + threadIdx.x; j <= colEndBound; j += blockDim.x)
for (auto c = 0; c < channels; c++) {
Nd4jLong zPos[] = {batch, rowStart, j, c};
Nd4jLong cPos[] = {colorIndex, c};
auto cIndex = shape::getOffset(colorTableShape, cPos, 0);
auto zIndex = shape::getOffset(outputShape, zPos, 0);
output[zIndex] = (T)colorTable[cIndex];
}
}
// Draw bottom line.
if (rowEnd < height) {
for (auto j = colStartBound + threadIdx.x; j <= colEndBound; j += blockDim.x)
for (auto c = 0; c < channels; c++) {
Nd4jLong zPos[] = {batch, rowEnd, j, c};
Nd4jLong cPos[] = {colorIndex, c};
auto cIndex = shape::getOffset(colorTableShape, cPos, 0);
auto zIndex = shape::getOffset(outputShape, zPos, 0);
output[zIndex] = (T)colorTable[cIndex];
}
}
// Draw left line.
if (colStart >= 0) {
for (auto i = rowStartBound + threadIdx.x; i <= rowEndBound; i += blockDim.x)
for (auto c = 0; c < channels; c++) {
Nd4jLong zPos[] = {batch, i, colStart, c};
Nd4jLong cPos[] = {colorIndex, c};
auto cIndex = shape::getOffset(colorTableShape, cPos, 0);
auto zIndex = shape::getOffset(outputShape, zPos, 0);
output[zIndex] = (T)colorTable[cIndex];
}
}
// Draw right line.
if (colEnd < width) {
for (auto i = rowStartBound + threadIdx.x; i <= rowEndBound; i += blockDim.x)
for (auto c = 0; c < channels; c++) {
Nd4jLong zPos[] = {batch, i, colEnd, c};
Nd4jLong cPos[] = {colorIndex, c};
auto cIndex = shape::getOffset(colorTableShape, cPos, 0);
auto zIndex = shape::getOffset(outputShape, zPos, 0);
output[zIndex] = (T)colorTable[cIndex];
}
}
}
}
@ -70,15 +143,19 @@ namespace helpers {
auto width = images->sizeAt(2);
auto channels = images->sizeAt(3);
auto stream = context->getCudaStream();
auto colorSetSize = colors->sizeAt(0);
auto boxSize = boxes->sizeAt(1);
NDArray colorsTable = DefaultColorTable(channels);
if ((colors != nullptr && colors->lengthOf() > 0)) {
colorsTable = *colors;
}
auto imagesBuf = images->getDataBuffer()->specialAsT<T>();
auto boxesBuf = boxes->getDataBuffer()->specialAsT<T>();
auto colorsBuf = colors->getDataBuffer()->specialAsT<T>();
auto boxesBuf = boxes->getDataBuffer()->specialAsT<float>(); // boxes should be float32
auto colorsTableBuf = colorsTable.getDataBuffer()->specialAsT<float>(); // color table is float32
auto outputBuf = output->dataBuffer()->specialAsT<T>();
drawBoundingBoxesKernel<<<batchSize > 128? 128: batchSize, 256, 1024, *stream>>>(imagesBuf, images->getSpecialShapeInfo(),
boxesBuf, boxes->getSpecialShapeInfo(), colorsBuf, colors->getSpecialShapeInfo(),
outputBuf, output->specialShapeInfo(), batchSize, width, height, channels, colorSetSize);
drawBoundingBoxesKernel<<<128, 128, 1024, *stream>>>(imagesBuf, images->getSpecialShapeInfo(),
boxesBuf, boxes->getSpecialShapeInfo(), colorsTableBuf, colorsTable.getSpecialShapeInfo(),
outputBuf, output->specialShapeInfo(), batchSize, width, height, channels, boxSize, colorsTable.lengthOf());
}
void drawBoundingBoxesFunctor(nd4j::LaunchContext * context, NDArray* images, NDArray* boxes, NDArray* colors, NDArray* output) {

207
libnd4j/tests_cpu/layers_tests/DeclarableOpsTests10.cpp
View File

@ -1470,6 +1470,64 @@ TEST_F(DeclarableOpsTests10, ImageResizeBilinear_Test1) {
delete results;
}
TEST_F(DeclarableOpsTests10, ImageResizeBilinear_Test01) {
NDArray input = NDArrayFactory::create<double>('c', {2,3,4});
//NDArray<float> paddings('c', {3,2}, {0,0, 0,1, 0,0});
//NDArray<float> expected('c', {2,4,4}, {1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,0.,0.,0.,0.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,0.,0.,0.,0.});
NDArray expected = NDArrayFactory::create<double>('c', {10, 10, 4}, {1., 2., 3., 4., 2.2, 3.2, 4.2, 5.2, 3.4, 4.4, 5.4, 6.4,
4.6, 5.6, 6.6, 7.6, 5.8, 6.8, 7.8, 8.8, 7., 8., 9., 10.,
8.2, 9.2, 10.2, 11.2, 9., 10., 11., 12., 9., 10., 11., 12.,
9., 10., 11., 12., 3.4, 4.4, 5.4, 6.4, 4.6, 5.6, 6.6, 7.6,
5.8, 6.8, 7.8, 8.8, 7.0, 8., 9., 10., 8.2, 9.2, 10.2, 11.2,
9.4,10.4, 11.4, 12.4,10.6, 11.6,12.6, 13.6,11.4, 12.4, 13.4, 14.4,
11.4,12.4, 13.4, 14.4,11.4, 12.4,13.4, 14.4, 5.8, 6.8, 7.8, 8.8,
7., 8., 9., 10., 8.2, 9.2,10.2, 11.2, 9.4, 10.4, 11.4, 12.4,
10.6,11.6, 12.6, 13.6,11.8, 12.8,13.8, 14.8,13.0, 14.0, 15.0, 16.,
13.8,14.8, 15.8, 16.8,13.8, 14.8,15.8, 16.8,13.8, 14.8, 15.8, 16.8,
8.2, 9.2, 10.2, 11.2, 9.4, 10.4,11.4, 12.4,10.6, 11.6, 12.6, 13.6,
11.8,12.8, 13.8, 14.8,13., 14., 15., 16., 14.2, 15.2, 16.2, 17.2,
15.4,16.4, 17.4, 18.4,16.2, 17.2,18.2, 19.2,16.2, 17.2, 18.2, 19.2,
16.2,17.2, 18.2, 19.2,10.6, 11.6,12.6, 13.6,11.8, 12.8, 13.8, 14.8,
13., 14., 15., 16., 14.2, 15.2,16.2, 17.2,15.4, 16.4, 17.4, 18.4,
16.6,17.6, 18.6, 19.6,17.8, 18.8,19.8, 20.8,18.6, 19.6, 20.6, 21.6,
18.6,19.6, 20.6, 21.6,18.6, 19.6,20.6, 21.6,13., 14., 15., 16.,
14.2,15.2, 16.2, 17.2,15.4, 16.4,17.4, 18.4,16.6, 17.6, 18.6, 19.6,
17.8,18.8, 19.8, 20.8,19., 20., 21., 22., 20.2, 21.2, 22.2, 23.2,
21., 22., 23., 24., 21., 22., 23., 24., 21., 22., 23., 24.,
13., 14., 15., 16., 14.2, 15.2,16.2, 17.2,15.4, 16.4, 17.4, 18.4,
16.6,17.6, 18.6, 19.6,17.8, 18.8, 19.8, 20.8,19., 20., 21., 22.,
20.2,21.2, 22.2, 23.2,21., 22., 23., 24., 21., 22., 23., 24.,
21., 22., 23., 24., 13., 14., 15., 16., 14.2, 15.2, 16.2, 17.2,
15.4,16.4, 17.4, 18.4,16.6, 17.6, 18.6, 19.6,17.8, 18.8, 19.8, 20.8,
19., 20., 21., 22., 20.2, 21.2, 22.2, 23.2,21., 22., 23., 24.,
21., 22., 23., 24., 21., 22., 23., 24., 13., 14., 15., 16.,
14.2,15.2, 16.2, 17.2,15.4, 16.4, 17.4, 18.4,16.6, 17.6, 18.6, 19.6,
17.8,18.8, 19.8, 20.8,19., 20., 21., 22., 20.2, 21.2, 22.2, 23.2,
21., 22., 23., 24., 21., 22., 23., 24., 21., 22., 23., 24.,
13., 14., 15., 16., 14.2, 15.2, 16.2, 17.2,15.4, 16.4, 17.4, 18.4,
16.6,17.6, 18.6, 19.6,17.8, 18.8, 19.8, 20.8,19., 20., 21., 22.,
20.2,21.2, 22.2, 23.2,
21., 22., 23., 24., 21., 22., 23., 24., 21., 22., 23., 24.});
//input = 1.f;
input.linspace(1);
nd4j::ops::resize_bilinear op;
auto results = op.execute({&input}, {}, {10, 10});
ASSERT_EQ(ND4J_STATUS_OK, results->status());
NDArray* result = results->at(0);
//result->printIndexedBuffer("Resized to 10x10");
//expected.printIndexedBuffer("Expect for 10x10");
// result->printShapeInfo("Output shape");
// expected.printShapeInfo("Expect shape");
ASSERT_TRUE(expected.isSameShape(result));
ASSERT_TRUE(expected.equalsTo(result));
delete results;
}
////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests10, ImageResizeBilinear_Test2) {
@ -1852,6 +1910,53 @@ TEST_F(DeclarableOpsTests10, ImageResizeNeighbor_Test1) {
delete results;
}
TEST_F(DeclarableOpsTests10, ImageResizeNeighbor_Test01) {
NDArray input = NDArrayFactory::create<double>('c', {2, 3, 4});
//NDArray<float> paddings('c', {3,2}, {0,0, 0,1, 0,0});
//NDArray<float> expected('c', {2,4,4}, {1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,0.,0.,0.,0.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,0.,0.,0.,0.});
NDArray expected = NDArrayFactory::create<double>('c', {4, 5, 4}, { 1, 2, 3, 4,
1, 2, 3, 4,
5, 6, 7, 8,
5, 6, 7, 8,
9, 10, 11, 12,
1, 2, 3, 4,
1, 2, 3, 4,
5, 6, 7, 8,
5, 6, 7, 8,
9, 10, 11, 12,
13, 14, 15, 16,
13, 14, 15, 16,
17, 18, 19, 20,
17, 18, 19, 20,
21, 22, 23, 24,
13, 14, 15, 16,
13, 14, 15, 16,
17, 18, 19, 20,
17, 18, 19, 20,
21, 22, 23, 24
});
//input = 1.f;
input.linspace(1);
nd4j::ops::resize_nearest_neighbor op;
auto results = op.execute({&input}, {}, {4, 5});
ASSERT_EQ(ND4J_STATUS_OK, results->status());
NDArray* result = results->at(0);
//result->printIndexedBuffer("Resized to 4x5");
//expected.printIndexedBuffer("Expect for 4x5");
ASSERT_TRUE(expected.isSameShape(result));
ASSERT_TRUE(expected.equalsTo(result));
delete results;
}
////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests10, ReduceLogSumExpTest_1) {
@ -2180,6 +2285,7 @@ TEST_F(DeclarableOpsTests10, Image_DrawBoundingBoxes_1) {
ASSERT_EQ(ND4J_STATUS_OK, results->status());
auto result = results->at(0);
result->syncToHost();
// result->printBuffer("Bounded boxes");
// expected.printBuffer("Bounded expec");
ASSERT_TRUE(expected.isSameShapeStrict(result));
@ -2221,6 +2327,58 @@ TEST_F(DeclarableOpsTests10, Image_DrawBoundingBoxes_2) {
delete results;
}
////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests10, Image_DrawBoundingBoxes_3) {
NDArray images = NDArrayFactory::create<float>('c', {2,5,5,1}, {0.7788, 0.8012, 0.7244, 0.2309, 0.7271, 0.1804,
0.5056, 0.8925, 0.5461, 0.9234, 0.0856,0.7938,
0.6591, 0.5555, 0.1596, 0.3087, 0.1548, 0.4695,
0.9939, 0.6113,0.6765, 0.1800, 0.6750, 0.2246,
0.0509, 0.4601, 0.8284, 0.2354, 0.9752, 0.8361,
0.2585, 0.4189, 0.7028, 0.7679, 0.5373, 0.7234,
0.2690, 0.0062, 0.0327, 0.0644, 0.8428, 0.7494,
0.0755, 0.6245, 0.3491, 0.5793, 0.5730, 0.1822,
0.6420, 0.9143});
NDArray boxes = NDArrayFactory::create<float>('c', {2, 2, 4}, {0.7717, 0.9281, 0.9846, 0.4838, 0.6433, 0.6041, 0.6501, 0.7612, 0.7605, 0.3948, 0.9493, 0.8600, 0.7876, 0.8945, 0.4638, 0.7157});
NDArray colors = NDArrayFactory::create<float>('c', {1, 2}, {0.9441, 0.5957});
//NDArray<float> ('c', {6}, {0.9f, .75f, .6f, .95f, .5f, .3f});
// NDArray expected = NDArrayFactory::create<float>('c', {2,5,5,1}, {
// 0.7788f, 0.8012f, 0.7244f, 0.2309f, 0.7271f,
// 0.1804f, 0.5056f, 0.8925f, 0.5461f, 0.9234f, 0.0856f, 0.7938f, 0.9441f,
// 0.9441f, 0.1596f, 0.3087f, 0.1548f, 0.4695f, 0.9939f, 0.6113f, 0.6765f,
// 0.1800f, 0.6750f, 0.2246f, 0.0509f, 0.4601f, 0.8284f, 0.2354f, 0.9752f, 0.8361f,
// 0.2585f, 0.4189f,0.7028f,0.7679f,0.5373f,0.7234f,0.2690f,0.0062f,0.0327f,0.0644f,
// 0.8428f, 0.9441f,0.9441f,0.9441f,0.3491f,0.5793f,0.5730f,0.1822f,0.6420f,0.9143f });
NDArray expected = NDArrayFactory::create<float>('c', {2,5,5,1}, {
0.7788, 0.8012, 0.7244, 0.2309, 0.7271,
0.1804, 0.5056, 0.8925, 0.5461, 0.9234,
0.0856, 0.7938, 0.9441, 0.9441, 0.1596,
0.3087, 0.1548, 0.4695, 0.9939, 0.6113,
0.6765, 0.18 , 0.675 , 0.2246, 0.0509,
0.4601, 0.8284, 0.2354, 0.9752, 0.8361,
0.2585, 0.4189, 0.7028, 0.7679, 0.5373,
0.7234, 0.269 , 0.0062, 0.0327, 0.0644,
0.8428, 0.9441, 0.9441, 0.9441, 0.3491,
0.5793, 0.573 , 0.1822, 0.642 , 0.9143});
nd4j::ops::draw_bounding_boxes op;
auto results = op.execute({&images, &boxes, &colors}, {}, {});
ASSERT_EQ(ND4J_STATUS_OK, results->status());
auto result = results->at(0);
// result->printBuffer("Boxes3 output");
// expected.printBuffer("Boxes3 expect");
// result->syncToHost();
// result->printBuffer("Bounded boxes 2");
// expected.printBuffer("Bounded expec 2");
ASSERT_TRUE(expected.isSameShapeStrict(result));
ASSERT_TRUE(expected.equalsTo(result));
delete results;
}
////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests10, FakeQuantWithMinMaxVars_Test_1) {
@ -2376,6 +2534,55 @@ TEST_F(DeclarableOpsTests10, FakeQuantWithMinMaxVars_Test_5) {
delete results;
}
/* public void testFakeQuantAgainstTF_1() {
INDArray x = Nd4j.createFromArray(new float[]{ 0.7788f,0.8012f, 0.7244f, 0.2309f,0.7271f,
0.1804f, 0.5056f, 0.8925f, 0.5461f, 0.9234f,
0.0856f, 0.7938f, 0.6591f, 0.5555f, 0.1596f}).reshape(3,5);
INDArray min = Nd4j.createFromArray(new float[]{-0.2283f, -0.0719f, -0.0154f, -0.5162f, -0.3567f}).reshape(1,5);
INDArray max = Nd4j.createFromArray(new float[]{0.9441f, 0.5957f, 0.8669f, 0.3502f, 0.5100f}).reshape(1,5);
INDArray out = Nd4j.createUninitialized(x.shape());
val op = new FakeQuantWithMinMaxVarsPerChannel(x,min,max,out);
INDArray expected = Nd4j.createFromArray(new float[]{0.7801f, 0.5966f, 0.7260f, 0.2320f, 0.5084f,
0.1800f, 0.5046f, 0.8684f, 0.3513f, 0.5084f,
0.0877f, 0.5966f, 0.6600f, 0.3513f, 0.1604f}).reshape(3,5);
assertEquals(expected, out);
}*/
TEST_F(DeclarableOpsTests10, FakeQuantWithMinMaxVars_Test_6) {
NDArray x = NDArrayFactory::create<float>('c', {3, 5}, {0.7788f,0.8012f, 0.7244f, 0.2309f,0.7271f,
0.1804f, 0.5056f, 0.8925f, 0.5461f, 0.9234f,
0.0856f, 0.7938f, 0.6591f, 0.5555f, 0.1596f});
// NDArray exp = NDArrayFactory::create<float>('c', {3, 5},{
// 0.7801f, 0.5966f, 0.7260f, 0.2320f, 0.5084f,
// 0.1800f, 0.5046f, 0.8684f, 0.3513f, 0.5084f,
// 0.0877f, 0.5966f, 0.6600f, 0.3513f, 0.1604f
// });
NDArray exp = NDArrayFactory::create<float>('c', {3,5}, {0.77700233, 0.596913, 0.72314, 0.23104, 0.50982356,
0.17930824, 0.50528157, 0.86846, 0.34995764, 0.50982356,
0.08735529, 0.596913, 0.6574, 0.34995764, 0.15974471});
NDArray min = NDArrayFactory::create<float>('c', {5}, {-0.2283f, -0.0719f, -0.0154f, -0.5162f, -0.3567f});
NDArray max = NDArrayFactory::create<float>('c', {5}, {0.9441f, 0.5957f, 0.8669f, 0.3502f, 0.5100f});
// x.linspace(-60.);
nd4j::ops::fake_quant_with_min_max_vars_per_channel op;
auto results = op.execute({&x, &min, &max}, {}, {});
ASSERT_EQ(ND4J_STATUS_OK, results->status());
auto result = results->at(0);
// result->printBuffer("Quantized per channels 5");
// exp.printBuffer("Quantized per channest E");
// auto diff = *result - exp;
// diff.printIndexedBuffer("Difference");
ASSERT_TRUE(exp.isSameShapeStrict(result));
ASSERT_TRUE(exp.equalsTo(result));
delete results;
}
////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests10, batchnorm_test1) {