* initial commit Signed-off-by: raver119 <raver119@gmail.com> * - gruCell_bp further Signed-off-by: Yurii <yurii@skymind.io> * - further work on gruCell_bp Signed-off-by: Yurii <yurii@skymind.io> * Inverse matrix cublas implementation. Partial working revision. * Separation of segment ops helpers. Max separation. * Separated segment_min ops. * Separation of segment_mean/sum/prod/sqrtN ops heleprs. * Fixed diagonal processing with LUP decomposition. * Modified inversion approach using current state of LU decomposition. * Implementation of matrix_inverse op with cuda kernels. Working revision. * Implemented sequence_mask cuda helper. Eliminated waste printf with matrix_inverse implementation. Added proper tests. * - further work on gruCell_bp (ff/cuda) Signed-off-by: Yurii <yurii@skymind.io> * comment one test for gruCell_bp Signed-off-by: Yurii <yurii@skymind.io> * - provide cuda static_rnn Signed-off-by: Yurii <yurii@skymind.io> * Refactored random_shuffle op to use new random generator. * Refactored random_shuffle op helper. * Fixed debug tests with random ops tests. * Implement random_shuffle op cuda kernel helper and tests. * - provide cuda scatter_update Signed-off-by: Yurii <yurii@skymind.io> * Implementation of random_shuffle for linear case with cuda kernels and tests. * Implemented random_shuffle with cuda kernels. Final revision. * - finally gruCell_bp is completed Signed-off-by: Yurii <yurii@skymind.io> * Dropout op cuda helper implementation. * Implemented dropout_bp cuda helper. * Implemented alpha_dropout_bp with cuda kernel helpers. * Refactored helper. * Implementation of suppresion helper with cuda kernels. * - provide cpu code fot hsvToRgb, rgbToHsv, adjustHue Signed-off-by: Yurii <yurii@skymind.io> * Using sort by value method. * Implementation of image.non_max_suppression op cuda-based helper. * - correcting and testing adjust_hue, adjust_saturation cpu/cuda code Signed-off-by: Yurii <yurii@skymind.io> * Added cuda device prefixes to declarations. * Implementation of hashcode op with cuda helper. Initital revision. * rnn cu impl removed Signed-off-by: raver119 <raver119@gmail.com>
121 lines
3.2 KiB
C++
121 lines
3.2 KiB
C++
/*******************************************************************************
|
|
* Copyright (c) 2015-2018 Skymind, Inc.
|
|
*
|
|
* This program and the accompanying materials are made available under the
|
|
* terms of the Apache License, Version 2.0 which is available at
|
|
* https://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.
|
|
*
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
******************************************************************************/
|
|
|
|
//
|
|
// @author raver119@gmail.com
|
|
// @author Yurii Shyrma (iuriish@yahoo.com)
|
|
//
|
|
|
|
#include <op_boilerplate.h>
|
|
#include <templatemath.h>
|
|
#include <NDArray.h>
|
|
|
|
namespace nd4j {
|
|
namespace ops {
|
|
namespace helpers {
|
|
|
|
void adjustSaturation(nd4j::LaunchContext* context, const NDArray *input, const NDArray* factorScalarArr, NDArray *output, const int dimC);
|
|
|
|
/*
|
|
template <typename T>
|
|
static FORCEINLINE _CUDA_HD void rgb_to_hsv(T r, T g, T b, T* h, T* s, T* v) {
|
|
T vv = nd4j::math::nd4j_max<T>(r, nd4j::math::nd4j_max<T>(g, b));
|
|
T range = vv - nd4j::math::nd4j_min<T>(r, nd4j::math::nd4j_min<T>(g, b));
|
|
if (vv > 0) {
|
|
*s = range / vv;
|
|
} else {
|
|
*s = 0;
|
|
}
|
|
T norm = 1.0f / (6.0f * range);
|
|
T hh;
|
|
if (r == vv) {
|
|
hh = norm * (g - b);
|
|
} else if (g == vv) {
|
|
hh = norm * (b - r) + 2.0 / 6.0;
|
|
} else {
|
|
hh = norm * (r - g) + 4.0 / 6.0;
|
|
}
|
|
if (range <= (T) 0.0) {
|
|
hh = 0;
|
|
}
|
|
if (hh < (T) 0.0) {
|
|
hh = hh + 1.;
|
|
}
|
|
*v = vv;
|
|
*h = hh;
|
|
}
|
|
|
|
template <typename T>
|
|
static FORCEINLINE _CUDA_HD void hsv_to_rgb(T h, T s, T v, T* r, T* g, T* b) {
|
|
T c = s * v;
|
|
T m = v - c;
|
|
T dh = h * 6;
|
|
T rr, gg, bb;
|
|
int h_category = static_cast<int>(dh);
|
|
T fmodu = dh;
|
|
while (fmodu <= (T) 0)
|
|
fmodu += (T) 2.0f;
|
|
|
|
while (fmodu >= (T) 2.0f)
|
|
fmodu -= (T) 2.0f;
|
|
|
|
T x = c * (1. - nd4j::math::nd4j_abs<T>(fmodu - 1.));
|
|
switch (h_category) {
|
|
case 0:
|
|
rr = c;
|
|
gg = x;
|
|
bb = 0;
|
|
break;
|
|
case 1:
|
|
rr = x;
|
|
gg = c;
|
|
bb = 0;
|
|
break;
|
|
case 2:
|
|
rr = 0;
|
|
gg = c;
|
|
bb = x;
|
|
break;
|
|
case 3:
|
|
rr = 0;
|
|
gg = x;
|
|
bb = c;
|
|
break;
|
|
case 4:
|
|
rr = x;
|
|
gg = 0;
|
|
bb = c;
|
|
break;
|
|
case 5:
|
|
rr = c;
|
|
gg = 0;
|
|
bb = x;
|
|
break;
|
|
default:
|
|
rr = 0;
|
|
gg = 0;
|
|
bb = 0;
|
|
}
|
|
|
|
*r = rr + m;
|
|
*g = gg + m;
|
|
*b = bb + m;
|
|
}
|
|
*/
|
|
|
|
}
|
|
}
|
|
} |