src/runtime/NEON/functions/NELocallyConnectedLayer.cpp

/*
 * Copyright (c) 2017 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 "arm_compute/runtime/NEON/functions/NELocallyConnectedLayer.h"

#include "arm_compute/core/PixelValue.h"
#include "arm_compute/core/Utils.h"
#include "arm_compute/core/Validate.h"
#include "arm_compute/runtime/NEON/NEScheduler.h"

#include <cmath>
#include <tuple>

using namespace arm_compute;

NELocallyConnectedLayer::NELocallyConnectedLayer(std::shared_ptr<IMemoryManager> memory_manager)
    : _memory_group(std::move(memory_manager)), _input_im2col_kernel(), _weights_reshape_kernel(), _mm_kernel(), _output_col2im_kernel(), _input_im2col_reshaped(), _weights_reshaped(), _gemm_output(),
      _is_first_run(false)
{
}

void NELocallyConnectedLayer::configure(const ITensor *input, const ITensor *weights, const ITensor *biases, ITensor *output, const PadStrideInfo &conv_info)
{
    ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input, 1, DataType::F32);
    ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(weights, 1, DataType::F32);
    ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output, 1, DataType::F32);
    ARM_COMPUTE_ERROR_ON_MISMATCHING_DATA_TYPES(input, weights, output);
    ARM_COMPUTE_ERROR_ON(weights->info()->dimension(2) != input->info()->dimension(2));

    if(biases != nullptr)
    {
        ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(biases, 1, DataType::F32);
        ARM_COMPUTE_ERROR_ON_MISMATCHING_DATA_TYPES(input, biases);
        ARM_COMPUTE_ERROR_ON(biases->info()->dimension(0) != weights->info()->dimension(3));
        ARM_COMPUTE_ERROR_ON(biases->info()->num_dimensions() > 2);
    }

    bool _has_bias = (biases != nullptr);
    _is_first_run  = true;

    // Get parameters for conv_info
    unsigned int stride_x = 0;
    unsigned int stride_y = 0;
    unsigned int pad_x    = 0;
    unsigned int pad_y    = 0;
    std::tie(stride_x, stride_y) = conv_info.stride();
    std::tie(pad_x, pad_y)       = conv_info.pad();

    const unsigned int kernel_width  = weights->info()->dimension(0);
    const unsigned int kernel_height = weights->info()->dimension(1);

    // Get convolved dimensions
    unsigned int conv_w = 0;
    unsigned int conv_h = 0;
    std::tie(conv_w, conv_h) = scaled_dimensions(input->info()->dimension(0), input->info()->dimension(1), kernel_width, kernel_height,
                                                 conv_info);

    ARM_COMPUTE_ERROR_ON_MSG((output->info()->dimension(0) != conv_w) || (output->info()->dimension(1) != conv_h), "Output shape does not match the expected one");
    ARM_COMPUTE_ERROR_ON_MSG(weights->info()->dimension(4) != (conv_w * conv_h), "Weights shape does not match the expected one");

    // Create tensor to store the reshaped weights
    const size_t mat_weights_cols = weights->info()->dimension(3);
    const size_t mat_weights_rows = weights->info()->dimension(0) * weights->info()->dimension(1) * weights->info()->dimension(2) + ((_has_bias) ? 1 : 0);
    const size_t mat_weights_num  = weights->info()->dimension(4);

    const TensorShape shape_wr(mat_weights_cols, mat_weights_rows, mat_weights_num);

    _weights_reshaped.allocator()->init(TensorInfo(shape_wr, 1, weights->info()->data_type()));

    // Create tensor to store im2col reshaped inputs
    const size_t mat_input_cols = mat_weights_rows;
    const size_t mat_input_rows = conv_w * conv_h;
    TensorShape  shape_im2col   = input->info()->tensor_shape();
    shape_im2col.set(0, mat_input_cols);
    shape_im2col.set(1, mat_input_rows);
    shape_im2col.set(2, 1);

    _input_im2col_reshaped.allocator()->init(TensorInfo(shape_im2col, 1, input->info()->data_type()));

    // Create locally connected layer output tensor
    TensorShape shape_gemm = _input_im2col_reshaped.info()->tensor_shape();
    shape_gemm.set(0, mat_weights_cols);
    shape_gemm.set(1, mat_input_rows);
    _gemm_output.allocator()->init(TensorInfo(shape_gemm, 1, input->info()->data_type()));

    // Manage intermediate buffers
    _memory_group.manage(&_input_im2col_reshaped);
    _memory_group.manage(&_gemm_output);

    // Configure kernels
    _input_im2col_kernel.configure(input, &_input_im2col_reshaped, Size2D(kernel_width, kernel_height), conv_info, _has_bias);
    _weights_reshape_kernel.configure(weights, biases, &_weights_reshaped);
    _mm_kernel.configure(&_input_im2col_reshaped, &_weights_reshaped, &_gemm_output);
    _output_col2im_kernel.configure(&_gemm_output, output, std::make_pair(conv_w, conv_h));

    // Allocate intermediate tensors
    _weights_reshaped.allocator()->allocate();
    _input_im2col_reshaped.allocator()->allocate();
    _gemm_output.allocator()->allocate();
}

void NELocallyConnectedLayer::run()
{
    // Run weights reshaping (Runs once for every configure)
    if(_is_first_run)
    {
        _is_first_run = false;
        NEScheduler::get().schedule(&_weights_reshape_kernel, 3);
    }

    _memory_group.acquire();

    // Run input reshaping
    NEScheduler::get().schedule(&_input_im2col_kernel, Window::DimY);

    // Runs GEMM on reshaped matrices
    NEScheduler::get().schedule(&_mm_kernel, Window::DimX);

    // Reshape output matrix
    NEScheduler::get().schedule(&_output_col2im_kernel, Window::DimY);

    _memory_group.release();
}