xref: /aosp_15_r20/external/ComputeLibrary/src/runtime/NEON/functions/NEDetectionPostProcessLayer.cpp (revision c217d954acce2dbc11938adb493fc0abd69584f3)
1 /*
2  * Copyright (c) 2019-2021 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/NEDetectionPostProcessLayer.h"
25 
26 #include "arm_compute/core/Error.h"
27 #include "arm_compute/core/Helpers.h"
28 #include "arm_compute/core/Validate.h"
29 #include "src/common/utils/Log.h"
30 
31 #include <cstddef>
32 #include <ios>
33 #include <list>
34 
35 namespace arm_compute
36 {
NEDetectionPostProcessLayer(std::shared_ptr<IMemoryManager> memory_manager)37 NEDetectionPostProcessLayer::NEDetectionPostProcessLayer(std::shared_ptr<IMemoryManager> memory_manager)
38     : _memory_group(std::move(memory_manager)), _dequantize(), _detection_post_process(), _decoded_scores(), _run_dequantize(false)
39 {
40 }
41 
configure(const ITensor * input_box_encoding,const ITensor * input_scores,const ITensor * input_anchors,ITensor * output_boxes,ITensor * output_classes,ITensor * output_scores,ITensor * num_detection,DetectionPostProcessLayerInfo info)42 void NEDetectionPostProcessLayer::configure(const ITensor *input_box_encoding, const ITensor *input_scores, const ITensor *input_anchors,
43                                             ITensor *output_boxes, ITensor *output_classes, ITensor *output_scores, ITensor *num_detection, DetectionPostProcessLayerInfo info)
44 {
45     ARM_COMPUTE_ERROR_ON_NULLPTR(input_box_encoding, input_scores, input_anchors, output_boxes, output_classes, output_scores);
46     ARM_COMPUTE_ERROR_THROW_ON(NEDetectionPostProcessLayer::validate(input_box_encoding->info(), input_scores->info(), input_anchors->info(), output_boxes->info(), output_classes->info(),
47                                                                      output_scores->info(),
48                                                                      num_detection->info(), info));
49     ARM_COMPUTE_LOG_PARAMS(input_box_encoding, input_scores, input_anchors, output_boxes, output_classes, output_scores, num_detection, info);
50 
51     const ITensor                *input_scores_to_use = input_scores;
52     DetectionPostProcessLayerInfo info_to_use         = info;
53     _run_dequantize                                   = is_data_type_quantized(input_box_encoding->info()->data_type());
54 
55     if(_run_dequantize)
56     {
57         _memory_group.manage(&_decoded_scores);
58 
59         _dequantize.configure(input_scores, &_decoded_scores);
60 
61         input_scores_to_use = &_decoded_scores;
62 
63         // Create a new info struct to avoid dequantizing in the CPP layer
64         std::array<float, 4> scales_values{ info.scale_value_y(), info.scale_value_x(), info.scale_value_h(), info.scale_value_w() };
65         DetectionPostProcessLayerInfo info_quantized(info.max_detections(), info.max_classes_per_detection(), info.nms_score_threshold(), info.iou_threshold(), info.num_classes(),
66                                                      scales_values, info.use_regular_nms(), info.detection_per_class(), false);
67         info_to_use = info_quantized;
68     }
69 
70     _detection_post_process.configure(input_box_encoding, input_scores_to_use, input_anchors, output_boxes, output_classes, output_scores, num_detection, info_to_use);
71     _decoded_scores.allocator()->allocate();
72 }
73 
validate(const ITensorInfo * input_box_encoding,const ITensorInfo * input_scores,const ITensorInfo * input_anchors,ITensorInfo * output_boxes,ITensorInfo * output_classes,ITensorInfo * output_scores,ITensorInfo * num_detection,DetectionPostProcessLayerInfo info)74 Status NEDetectionPostProcessLayer::validate(const ITensorInfo *input_box_encoding, const ITensorInfo *input_scores, const ITensorInfo *input_anchors,
75                                              ITensorInfo *output_boxes, ITensorInfo *output_classes, ITensorInfo *output_scores, ITensorInfo *num_detection, DetectionPostProcessLayerInfo info)
76 {
77     bool run_dequantize = is_data_type_quantized(input_box_encoding->data_type());
78     if(run_dequantize)
79     {
80         TensorInfo decoded_classes_info = input_scores->clone()->set_is_resizable(true).set_data_type(DataType::F32);
81         ARM_COMPUTE_RETURN_ON_ERROR(NEDequantizationLayer::validate(input_scores, &decoded_classes_info));
82     }
83     ARM_COMPUTE_RETURN_ON_ERROR(CPPDetectionPostProcessLayer::validate(input_box_encoding, input_scores, input_anchors, output_boxes, output_classes, output_scores, num_detection, info));
84 
85     return Status{};
86 }
87 
run()88 void NEDetectionPostProcessLayer::run()
89 {
90     MemoryGroupResourceScope scope_mg(_memory_group);
91 
92     // Decode scores if necessary
93     if(_run_dequantize)
94     {
95         _dequantize.run();
96     }
97     _detection_post_process.run();
98 }
99 } // namespace arm_compute
100