1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
|
/*
* Copyright (c) 2017-2020 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to
* deal in the Software without restriction, including without limitation the
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
* sell copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#include "Accumulate.h"
#include "arm_compute/core/Types.h"
#include "tests/validation/Helpers.h"
namespace arm_compute
{
namespace test
{
namespace validation
{
namespace reference
{
template <typename T1, typename T2>
SimpleTensor<T2> accumulate(const SimpleTensor<T1> &src, DataType output_data_type)
{
SimpleTensor<T2> dst{ src.shape(), output_data_type };
library->fill_tensor_uniform(dst, 1, static_cast<T2>(0), static_cast<T2>(std::numeric_limits<T1>::max()));
using intermediate_type = typename common_promoted_signed_type<T1, T2>::intermediate_type;
#if defined(_OPENMP)
#pragma omp parallel for
#endif /* _OPENMP */
for(int i = 0; i < src.num_elements(); ++i)
{
intermediate_type val = static_cast<intermediate_type>(src[i]) + static_cast<intermediate_type>(dst[i]);
dst[i] = saturate_cast<T2>(val);
}
return dst;
}
template <typename T1, typename T2>
SimpleTensor<T2> accumulate_weighted(const SimpleTensor<T1> &src, float alpha, DataType output_data_type)
{
ARM_COMPUTE_ERROR_ON_MSG(alpha < 0.f || alpha > 1.f, "Weight (alpha) specified in accumulate_weighted must be within the range [0, 1]");
SimpleTensor<T2> dst{ src.shape(), output_data_type };
library->fill_tensor_uniform(dst, 1, static_cast<T2>(0), static_cast<T2>(std::numeric_limits<T1>::max()));
using intermediate_type = typename common_promoted_signed_type<T1, T2>::intermediate_type;
#if defined(_OPENMP)
#pragma omp parallel for
#endif /* _OPENMP */
for(int i = 0; i < src.num_elements(); ++i)
{
double val = (1. - static_cast<double>(alpha)) * static_cast<intermediate_type>(dst[i]) + static_cast<double>(alpha) * static_cast<intermediate_type>(src[i]);
dst[i] = static_cast<T2>(val);
}
return dst;
}
template <typename T1, typename T2>
SimpleTensor<T2> accumulate_squared(const SimpleTensor<T1> &src, uint32_t shift, DataType output_data_type)
{
ARM_COMPUTE_ERROR_ON_MSG(shift > 15, "Shift in accumulate_squared must be within the range [0, 15]");
SimpleTensor<T2> dst{ src.shape(), output_data_type };
library->fill_tensor_uniform(dst, 1, static_cast<T2>(0), static_cast<T2>(std::numeric_limits<T1>::max()));
using intermediate_type = typename common_promoted_signed_type<T1, T2>::intermediate_type;
intermediate_type denom = 1 << shift;
#if defined(_OPENMP)
#pragma omp parallel for
#endif /* _OPENMP */
for(int i = 0; i < src.num_elements(); ++i)
{
intermediate_type val = static_cast<intermediate_type>(dst[i]) + (static_cast<intermediate_type>(src[i]) * static_cast<intermediate_type>(src[i]) / denom);
dst[i] = saturate_cast<T2>(val);
}
return dst;
}
template SimpleTensor<int16_t> accumulate(const SimpleTensor<uint8_t> &src, DataType output_data_type);
template SimpleTensor<uint8_t> accumulate_weighted(const SimpleTensor<uint8_t> &src, float alpha, DataType output_data_type);
template SimpleTensor<int16_t> accumulate_squared(const SimpleTensor<uint8_t> &src, uint32_t shift, DataType output_data_type);
} // namespace reference
} // namespace validation
} // namespace test
} // namespace arm_compute
|
