darknet/src/col2im_kernels.cu

115 lines
3.8 KiB
Plaintext
Raw Normal View History

2015-01-23 03:38:24 +03:00
extern "C" {
#include "col2im.h"
#include "cuda.h"
}
// src: https://github.com/BVLC/caffe/blob/master/src/caffe/util/im2col.cu
// You may also want to read: https://github.com/BVLC/caffe/blob/master/LICENSE
2014-08-28 06:11:46 +04:00
__global__ void col2im_gpu_kernel(const int n, const float* data_col,
const int height, const int width, const int ksize,
const int pad,
const int stride,
const int height_col, const int width_col,
float *data_im) {
int index = blockIdx.x*blockDim.x+threadIdx.x;
for(; index < n; index += blockDim.x*gridDim.x){
float val = 0;
int w = index % width + pad;
int h = (index / width) % height + pad;
int c = index / (width * height);
// compute the start and end of the output
int w_col_start = (w < ksize) ? 0 : (w - ksize) / stride + 1;
int w_col_end = min(w / stride + 1, width_col);
int h_col_start = (h < ksize) ? 0 : (h - ksize) / stride + 1;
int h_col_end = min(h / stride + 1, height_col);
// equivalent implementation
int offset =
(c * ksize * ksize + h * ksize + w) * height_col * width_col;
int coeff_h_col = (1 - stride * ksize * height_col) * width_col;
int coeff_w_col = (1 - stride * height_col * width_col);
for (int h_col = h_col_start; h_col < h_col_end; ++h_col) {
for (int w_col = w_col_start; w_col < w_col_end; ++w_col) {
val += data_col[offset + h_col * coeff_h_col + w_col * coeff_w_col];
}
2014-08-28 06:11:46 +04:00
}
data_im[index] = val;
2014-08-28 06:11:46 +04:00
}
}
2015-03-27 05:13:59 +03:00
void col2im_ongpu(float *data_col,
int channels, int height, int width,
2015-03-27 05:13:59 +03:00
int ksize, int stride, int pad, float *data_im){
// We are going to launch channels * height_col * width_col kernels, each
// kernel responsible for copying a single-channel grid.
pad = pad ? ksize/2 : 0;
int height_col = (height + 2 * pad - ksize) / stride + 1;
int width_col = (width + 2 * pad - ksize) / stride + 1;
int num_kernels = channels * height * width;
col2im_gpu_kernel<<<(num_kernels+BLOCK-1)/BLOCK,
BLOCK>>>(
num_kernels, data_col, height, width, ksize, pad,
stride, height_col,
2015-03-27 05:13:59 +03:00
width_col, data_im);
}
/*
__global__ void col2im_kernel(float *data_col,
int channels, int height, int width,
int ksize, int stride, int pad, float *data_im)
{
int height_col = (height - ksize) / stride + 1;
int width_col = (width - ksize) / stride + 1;
if (pad){
height_col = 1 + (height-1) / stride;
width_col = 1 + (width-1) / stride;
pad = ksize/2;
}
int id = (blockIdx.x + blockIdx.y*gridDim.x) * blockDim.x + threadIdx.x;
if(id >= channels*height*width) return;
int index = id;
int w = id%width + pad;
id /= width;
int h = id%height + pad;
id /= height;
int c = id%channels;
int w_start = (w-ksize+stride)/stride;
int w_end = w/stride + 1;
int h_start = (h-ksize+stride)/stride;
int h_end = h/stride + 1;
// int rows = channels * ksize * ksize;
// int cols = height_col*width_col;
int col_offset = (c*ksize*ksize + h * ksize + w)*height_col*width_col;
int h_coeff = (1-stride*ksize*height_col)*width_col;
int w_coeff = 1-stride*height_col*width_col;
float val = 0;
int h_col, w_col;
for(h_col = h_start; h_col < h_end; ++h_col){
for(w_col = w_start; w_col < w_end; ++w_col){
int col_index = col_offset +h_col*h_coeff + w_col*w_coeff;
float part = (w_col < 0 || h_col < 0 || h_col >= height_col || w_col >= width_col) ? 0 : data_col[col_index];
val += part;
}
}
data_im[index] = val;
2014-08-28 06:11:46 +04:00
}
2015-01-23 03:38:24 +03:00
2015-02-11 06:41:03 +03:00
extern "C" void col2im_ongpu(float *data_col,
int channels, int height, int width,
int ksize, int stride, int pad, float *data_im)
2015-01-23 03:38:24 +03:00
{
size_t n = channels*height*width;
2015-01-23 03:38:24 +03:00
col2im_kernel<<<cuda_gridsize(n), BLOCK>>>(data_col, channels, height, width, ksize, stride, pad, data_im);
check_error(cudaPeekAtLastError());
2015-01-23 03:38:24 +03:00
}
*/