1 /*
2 * Copyright (c) 2017-2021 Arm Limited.
3 *
4 * SPDX-License-Identifier: MIT
5 *
6 * Permission is hereby granted, free of charge, to any person obtaining a copy
7 * of this software and associated documentation files (the "Software"), to
8 * deal in the Software without restriction, including without limitation the
9 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
10 * sell copies of the Software, and to permit persons to whom the Software is
11 * furnished to do so, subject to the following conditions:
12 *
13 * The above copyright notice and this permission notice shall be included in all
14 * copies or substantial portions of the Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
22 * SOFTWARE.
23 */
24 #include "arm_compute/runtime/NEON/functions/NEL2NormalizeLayer.h"
25
26 #include "arm_compute/core/Helpers.h"
27 #include "arm_compute/runtime/NEON/NEScheduler.h"
28 #include "src/common/utils/Log.h"
29 #include "src/core/NEON/kernels/NEL2NormalizeLayerKernel.h"
30 #include "src/core/NEON/kernels/NEReductionOperationKernel.h"
31
32 namespace arm_compute
33 {
34 namespace
35 {
36 constexpr int max_input_tensor_dim = 3;
37 } // namespace
38 NEL2NormalizeLayer::~NEL2NormalizeLayer() = default;
39
NEL2NormalizeLayer(std::shared_ptr<IMemoryManager> memory_manager)40 NEL2NormalizeLayer::NEL2NormalizeLayer(std::shared_ptr<IMemoryManager> memory_manager)
41 : _memory_group(std::move(memory_manager)), _reduce_func(), _normalize_kernel(), _sumsq()
42 {
43 }
44
configure(ITensor * input,ITensor * output,int axis,float epsilon)45 void NEL2NormalizeLayer::configure(ITensor *input, ITensor *output, int axis, float epsilon)
46 {
47 ARM_COMPUTE_LOG_PARAMS(input, output, axis, epsilon);
48
49 // Manage intermediate buffers
50 _memory_group.manage(&_sumsq);
51
52 // Configure Kernels
53 const uint32_t actual_axis = wrap_around(axis, max_input_tensor_dim);
54 _reduce_func.configure(input, &_sumsq, actual_axis, ReductionOperation::SUM_SQUARE);
55 _normalize_kernel = std::make_unique<NEL2NormalizeLayerKernel>();
56 _normalize_kernel->configure(input, &_sumsq, output, axis, epsilon);
57
58 // Allocate intermediate tensors
59 _sumsq.allocator()->allocate();
60 }
61
validate(const ITensorInfo * input,const ITensorInfo * output,int axis,float epsilon)62 Status NEL2NormalizeLayer::validate(const ITensorInfo *input, const ITensorInfo *output, int axis, float epsilon)
63 {
64 TensorShape shape(input->tensor_shape());
65
66 // Create intermediate tensor info
67 TensorInfo sum_sq;
68 sum_sq.set_data_type(input->data_type());
69 sum_sq.set_tensor_shape(shape);
70
71 const uint32_t actual_axis = wrap_around(axis, max_input_tensor_dim);
72 ARM_COMPUTE_RETURN_ON_ERROR(NEReductionOperation::validate(input, &sum_sq, actual_axis, ReductionOperation::SUM_SQUARE));
73
74 // Reduce shape on axis
75 shape.set(actual_axis, 1);
76 sum_sq.set_tensor_shape(shape);
77
78 ARM_COMPUTE_RETURN_ON_ERROR(NEL2NormalizeLayerKernel::validate(input, &sum_sq, output, axis, epsilon));
79
80 return Status{};
81 }
82
run()83 void NEL2NormalizeLayer::run()
84 {
85 MemoryGroupResourceScope scope_mg(_memory_group);
86
87 _reduce_func.run();
88 NEScheduler::get().schedule(_normalize_kernel.get(), Window::DimY);
89 }
90 } // namespace arm_compute
91