github
/
darknet
mirror of https://github.com/AlexeyAB/darknet.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Use ignore_thresh only if class_id matched. Temporary changed Assisted_Excitation (reduces background activations rather than enhancing objects activations). Added antialiasiong=2 for 2x2.

Browse Source
pull/4140/head
AlexeyAB 6 years ago
parent 2eb68d5177
commit e6486ab594
6 changed files with 151 additions and 65 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 40
      src/convolutional_kernels.cu
  2. 70
      src/convolutional_layer.c
  3. 3
      src/detector.c
  4. 2
      src/http_stream.cpp
  5. 69
      src/maxpool_layer.c
  6. 32
      src/yolo_layer.c

40
src/convolutional_kernels.cu
Unescape View File

@ -948,6 +948,30 @@ void assisted_activation_gpu(float alpha, float *output, float *gt_gpu, float *a
assisted_activation_kernel << <num_blocks, BLOCK, 0, get_cuda_stream() >> > (alpha, output, gt_gpu, a_avg_gpu, size, channels, batches); assisted_activation_kernel << <num_blocks, BLOCK, 0, get_cuda_stream() >> > (alpha, output, gt_gpu, a_avg_gpu, size, channels, batches);
} }
__global__ void assisted_activation2_kernel(float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
{
int i = blockIdx.x * blockDim.x + threadIdx.x;
int xy = i % size;
int b = i / size;
float beta = 1 - alpha;
if (b < batches) {
for (int c = 0; c < channels; ++c) {
if(gt_gpu[i] == 0)
output[xy + size*(c + channels*b)] *= beta;
}
}
}
void assisted_activation2_gpu(float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
{
const int num_blocks = get_number_of_blocks(size*batches, BLOCK);
assisted_activation2_kernel << <num_blocks, BLOCK, 0, get_cuda_stream() >> > (alpha, output, gt_gpu, a_avg_gpu, size, channels, batches);
}
void assisted_excitation_forward_gpu(convolutional_layer l, network_state state) void assisted_excitation_forward_gpu(convolutional_layer l, network_state state)
{ {
const int iteration_num = (*state.net.seen) / (state.net.batch*state.net.subdivisions); const int iteration_num = (*state.net.seen) / (state.net.batch*state.net.subdivisions);
@ -958,12 +982,13 @@ void assisted_excitation_forward_gpu(convolutional_layer l, network_state state)
// calculate alpha // calculate alpha
//const float alpha = (1 + cos(3.141592 * iteration_num)) / (2 * state.net.max_batches); //const float alpha = (1 + cos(3.141592 * iteration_num)) / (2 * state.net.max_batches);
//const float alpha = (1 + cos(3.141592 * epoch)) / (2 * state.net.max_batches); //const float alpha = (1 + cos(3.141592 * epoch)) / (2 * state.net.max_batches);
//const float alpha = (1 + cos(3.141592 * iteration_num / state.net.max_batches)) / 2; float alpha = (1 + cos(3.141592 * iteration_num / state.net.max_batches)) / 2;
float alpha = (1 + cos(3.141592 * iteration_num / state.net.max_batches)); //float alpha = (1 + cos(3.141592 * iteration_num / state.net.max_batches));
if (l.assisted_excitation > 1) { if (l.assisted_excitation > 1) {
if (iteration_num > l.assisted_excitation) alpha = 0; if (iteration_num < state.net.burn_in) alpha = 0;
else alpha = (1 + cos(3.141592 * iteration_num / l.assisted_excitation)); else if (iteration_num > l.assisted_excitation) alpha = 0;
else alpha = (1 + cos(3.141592 * iteration_num / l.assisted_excitation)) / 2;
} }
//printf("\n epoch = %f, alpha = %f, seen = %d, max_batches = %d, train_images_num = %d \n", //printf("\n epoch = %f, alpha = %f, seen = %d, max_batches = %d, train_images_num = %d \n",
@ -1017,7 +1042,8 @@ void assisted_excitation_forward_gpu(convolutional_layer l, network_state state)
//CHECK_CUDA(cudaPeekAtLastError()); //CHECK_CUDA(cudaPeekAtLastError());
// calc new output // calc new output
assisted_activation_gpu(alpha, l.output_gpu, l.gt_gpu, l.a_avg_gpu, l.out_w * l.out_h, l.out_c, l.batch); assisted_activation2_gpu(alpha, l.output_gpu, l.gt_gpu, l.a_avg_gpu, l.out_w * l.out_h, l.out_c, l.batch);
//assisted_activation_gpu(alpha, l.output_gpu, l.gt_gpu, l.a_avg_gpu, l.out_w * l.out_h, l.out_c, l.batch);
//cudaStreamSynchronize(get_cuda_stream()); //cudaStreamSynchronize(get_cuda_stream());
//CHECK_CUDA(cudaPeekAtLastError()); //CHECK_CUDA(cudaPeekAtLastError());
@ -1070,13 +1096,13 @@ void assisted_excitation_forward_gpu(convolutional_layer l, network_state state)
printf(" Assisted Excitation alpha = %f \n", alpha); printf(" Assisted Excitation alpha = %f \n", alpha);
image img = float_to_image(l.out_w, l.out_h, 1, &gt[l.out_w*l.out_h*b]); image img = float_to_image(l.out_w, l.out_h, 1, &gt[l.out_w*l.out_h*b]);
char buff[100]; char buff[100];
sprintf(buff, "a_excitation_%d", b); sprintf(buff, "a_excitation_gt_%d", b);
show_image_cv(img, buff); show_image_cv(img, buff);
//image img2 = float_to_image(l.out_w, l.out_h, 1, &l.output[l.out_w*l.out_h*l.out_c*b]); //image img2 = float_to_image(l.out_w, l.out_h, 1, &l.output[l.out_w*l.out_h*l.out_c*b]);
image img2 = float_to_image_scaled(l.out_w, l.out_h, 1, &l.output[l.out_w*l.out_h*l.out_c*b]); image img2 = float_to_image_scaled(l.out_w, l.out_h, 1, &l.output[l.out_w*l.out_h*l.out_c*b]);
char buff2[100]; char buff2[100];
sprintf(buff2, "a_excitation_act_%d", b); sprintf(buff2, "a_excitation_output_%d", b);
show_image_cv(img2, buff2); show_image_cv(img2, buff2);
/* /*

70
src/convolutional_layer.c
Unescape View File

@ -587,36 +587,50 @@ convolutional_layer make_convolutional_layer(int batch, int steps, int h, int w,
if (l.antialiasing) { if (l.antialiasing) {
printf("AA: "); printf("AA: ");
l.input_layer = (layer*)calloc(1, sizeof(layer)); l.input_layer = (layer*)calloc(1, sizeof(layer));
const int blur_size = 3; int blur_size = 3;
*(l.input_layer) = make_convolutional_layer(batch, steps, out_h, out_w, n, n, n, blur_size, blur_stride_x, blur_stride_y, 1, blur_size / 2, LINEAR, 0, 0, 0, 0, 0, index, 0, NULL, 0); int blur_pad = blur_size / 2;
if (l.antialiasing == 2) {
blur_size = 2;
blur_pad = 0;
}
*(l.input_layer) = make_convolutional_layer(batch, steps, out_h, out_w, n, n, n, blur_size, blur_stride_x, blur_stride_y, 1, blur_pad, LINEAR, 0, 0, 0, 0, 0, index, 0, NULL, 0);
const int blur_nweights = n * blur_size * blur_size; // (n / n) * n * blur_size * blur_size; const int blur_nweights = n * blur_size * blur_size; // (n / n) * n * blur_size * blur_size;
int i; int i;
for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) { if (blur_size == 2) {
/* for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) {
l.input_layer->weights[i + 0] = 0; l.input_layer->weights[i + 0] = 1 / 4.f;
l.input_layer->weights[i + 1] = 0; l.input_layer->weights[i + 1] = 1 / 4.f;
l.input_layer->weights[i + 2] = 0; l.input_layer->weights[i + 2] = 1 / 4.f;
l.input_layer->weights[i + 3] = 1 / 4.f;
l.input_layer->weights[i + 3] = 0; }
l.input_layer->weights[i + 4] = 1; }
l.input_layer->weights[i + 5] = 0; else {
for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) {
l.input_layer->weights[i + 6] = 0; /*
l.input_layer->weights[i + 7] = 0; l.input_layer->weights[i + 0] = 0;
l.input_layer->weights[i + 8] = 0; l.input_layer->weights[i + 1] = 0;
*/ l.input_layer->weights[i + 2] = 0;
l.input_layer->weights[i + 0] = 1 / 16.f;
l.input_layer->weights[i + 1] = 2 / 16.f; l.input_layer->weights[i + 3] = 0;
l.input_layer->weights[i + 2] = 1 / 16.f; l.input_layer->weights[i + 4] = 1;
l.input_layer->weights[i + 5] = 0;
l.input_layer->weights[i + 3] = 2 / 16.f;
l.input_layer->weights[i + 4] = 4 / 16.f; l.input_layer->weights[i + 6] = 0;
l.input_layer->weights[i + 5] = 2 / 16.f; l.input_layer->weights[i + 7] = 0;
l.input_layer->weights[i + 8] = 0;
l.input_layer->weights[i + 6] = 1 / 16.f; */
l.input_layer->weights[i + 7] = 2 / 16.f; l.input_layer->weights[i + 0] = 1 / 16.f;
l.input_layer->weights[i + 8] = 1 / 16.f; l.input_layer->weights[i + 1] = 2 / 16.f;
l.input_layer->weights[i + 2] = 1 / 16.f;
l.input_layer->weights[i + 3] = 2 / 16.f;
l.input_layer->weights[i + 4] = 4 / 16.f;
l.input_layer->weights[i + 5] = 2 / 16.f;
l.input_layer->weights[i + 6] = 1 / 16.f;
l.input_layer->weights[i + 7] = 2 / 16.f;
l.input_layer->weights[i + 8] = 1 / 16.f;
}
} }
for (i = 0; i < n; ++i) l.input_layer->biases[i] = 0; for (i = 0; i < n; ++i) l.input_layer->biases[i] = 0;
#ifdef GPU #ifdef GPU

3
src/detector.c
Unescape View File

@ -798,7 +798,7 @@ float validate_detector_map(char *datacfg, char *cfgfile, char *weightfile, floa
replace_image_to_label(path, labelpath); replace_image_to_label(path, labelpath);
int num_labels = 0; int num_labels = 0;
box_label *truth = read_boxes(labelpath, &num_labels); box_label *truth = read_boxes(labelpath, &num_labels);
int i, j; int j;
for (j = 0; j < num_labels; ++j) { for (j = 0; j < num_labels; ++j) {
truth_classes_count[truth[j].id]++; truth_classes_count[truth[j].id]++;
} }
@ -818,6 +818,7 @@ float validate_detector_map(char *datacfg, char *cfgfile, char *weightfile, floa
const int checkpoint_detections_count = detections_count; const int checkpoint_detections_count = detections_count;
int i;
for (i = 0; i < nboxes; ++i) { for (i = 0; i < nboxes; ++i) {
int class_id; int class_id;

2
src/http_stream.cpp
Unescape View File

@ -48,7 +48,7 @@ static int close_socket(SOCKET s) {
cerr << "Close socket: out = " << close_output << ", in = " << close_input << " \n"; cerr << "Close socket: out = " << close_output << ", in = " << close_input << " \n";
return result; return result;
} }
#else // nix #else // _WIN32 - else: nix
#include "darkunistd.h" #include "darkunistd.h"
#include <sys/time.h> #include <sys/time.h>
#include <sys/types.h> #include <sys/types.h>

69
src/maxpool_layer.c
Unescape View File

@ -108,35 +108,50 @@ maxpool_layer make_maxpool_layer(int batch, int h, int w, int c, int size, int s
if (l.antialiasing) { if (l.antialiasing) {
printf("AA: "); printf("AA: ");
l.input_layer = (layer*)calloc(1, sizeof(layer)); l.input_layer = (layer*)calloc(1, sizeof(layer));
const int blur_size = 3; int blur_size = 3;
*(l.input_layer) = make_convolutional_layer(batch, 1, l.out_h, l.out_w, l.out_c, l.out_c, l.out_c, blur_size, blur_stride_x, blur_stride_y, 1, blur_size / 2, LINEAR, 0, 0, 0, 0, 0, 1, 0, NULL, 0); int blur_pad = blur_size / 2;
if (l.antialiasing == 2) {
blur_size = 2;
blur_pad = 0;
}
*(l.input_layer) = make_convolutional_layer(batch, 1, l.out_h, l.out_w, l.out_c, l.out_c, l.out_c, blur_size, blur_stride_x, blur_stride_y, 1, blur_pad, LINEAR, 0, 0, 0, 0, 0, 1, 0, NULL, 0);
const int blur_nweights = l.out_c * blur_size * blur_size; // (n / n) * n * blur_size * blur_size; const int blur_nweights = l.out_c * blur_size * blur_size; // (n / n) * n * blur_size * blur_size;
int i; int i;
for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) { if (blur_size == 2) {
/* for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) {
l.input_layer->weights[i + 0] = 0; l.input_layer->weights[i + 0] = 1 / 4.f;
l.input_layer->weights[i + 1] = 0; l.input_layer->weights[i + 1] = 1 / 4.f;
l.input_layer->weights[i + 2] = 0; l.input_layer->weights[i + 2] = 1 / 4.f;
l.input_layer->weights[i + 3] = 1 / 4.f;
l.input_layer->weights[i + 3] = 0; }
l.input_layer->weights[i + 4] = 1; }
l.input_layer->weights[i + 5] = 0; else {
for (i = 0; i < blur_nweights; i += (blur_size*blur_size)) {
l.input_layer->weights[i + 6] = 0; /*
l.input_layer->weights[i + 7] = 0; l.input_layer->weights[i + 0] = 0;
l.input_layer->weights[i + 8] = 0; l.input_layer->weights[i + 1] = 0;
*/ l.input_layer->weights[i + 2] = 0;
l.input_layer->weights[i + 0] = 1 / 16.f;
l.input_layer->weights[i + 1] = 2 / 16.f; l.input_layer->weights[i + 3] = 0;
l.input_layer->weights[i + 2] = 1 / 16.f; l.input_layer->weights[i + 4] = 1;
l.input_layer->weights[i + 5] = 0;
l.input_layer->weights[i + 3] = 2 / 16.f;
l.input_layer->weights[i + 4] = 4 / 16.f; l.input_layer->weights[i + 6] = 0;
l.input_layer->weights[i + 5] = 2 / 16.f; l.input_layer->weights[i + 7] = 0;
l.input_layer->weights[i + 8] = 0;
l.input_layer->weights[i + 6] = 1 / 16.f; */
l.input_layer->weights[i + 7] = 2 / 16.f; l.input_layer->weights[i + 0] = 1 / 16.f;
l.input_layer->weights[i + 8] = 1 / 16.f; l.input_layer->weights[i + 1] = 2 / 16.f;
l.input_layer->weights[i + 2] = 1 / 16.f;
l.input_layer->weights[i + 3] = 2 / 16.f;
l.input_layer->weights[i + 4] = 4 / 16.f;
l.input_layer->weights[i + 5] = 2 / 16.f;
l.input_layer->weights[i + 6] = 1 / 16.f;
l.input_layer->weights[i + 7] = 2 / 16.f;
l.input_layer->weights[i + 8] = 1 / 16.f;
}
} }
for (i = 0; i < l.out_c; ++i) l.input_layer->biases[i] = 0; for (i = 0; i < l.out_c; ++i) l.input_layer->biases[i] = 0;
#ifdef GPU #ifdef GPU

32
src/yolo_layer.c
Unescape View File

@ -128,6 +128,26 @@ box get_yolo_box(float *x, float *biases, int n, int index, int i, int j, int lw
return b; return b;
} }
int get_yolo_class(float *output, int classes, int class_index, int stride, float objectness)
{
int class_id = 0;
float max_prob = FLT_MIN;
int j;
for (j = 0; j < classes; ++j) {
float prob = objectness * output[class_index + stride*j];
if (prob > max_prob) {
max_prob = prob;
class_id = j;
}
//int class_index = entry_index(l, 0, n*l.w*l.h + i, 4 + 1 + j);
//float prob = objectness*predictions[class_index];
//dets[count].prob[j] = (prob > thresh) ? prob : 0;
}
return class_id;
}
ious delta_yolo_box(box truth, float *x, float *biases, int n, int index, int i, int j, int lw, int lh, int w, int h, float *delta, float scale, int stride, float iou_normalizer, IOU_LOSS iou_loss) ious delta_yolo_box(box truth, float *x, float *biases, int n, int index, int i, int j, int lw, int lh, int w, int h, float *delta, float scale, int stride, float iou_normalizer, IOU_LOSS iou_loss)
{ {
ious all_ious = { 0 }; ious all_ious = { 0 };
@ -272,6 +292,7 @@ void forward_yolo_layer(const layer l, network_state state)
box pred = get_yolo_box(l.output, l.biases, l.mask[n], box_index, i, j, l.w, l.h, state.net.w, state.net.h, l.w*l.h); box pred = get_yolo_box(l.output, l.biases, l.mask[n], box_index, i, j, l.w, l.h, state.net.w, state.net.h, l.w*l.h);
float best_iou = 0; float best_iou = 0;
int best_t = 0; int best_t = 0;
int class_id_match = 0;
for (t = 0; t < l.max_boxes; ++t) { for (t = 0; t < l.max_boxes; ++t) {
box truth = float_to_box_stride(state.truth + t*(4 + 1) + b*l.truths, 1); box truth = float_to_box_stride(state.truth + t*(4 + 1) + b*l.truths, 1);
int class_id = state.truth[t*(4 + 1) + b*l.truths + 4]; int class_id = state.truth[t*(4 + 1) + b*l.truths + 4];
@ -282,8 +303,17 @@ void forward_yolo_layer(const layer l, network_state state)
continue; // if label contains class_id more than number of classes in the cfg-file continue; // if label contains class_id more than number of classes in the cfg-file
} }
if (!truth.x) break; // continue; if (!truth.x) break; // continue;
int class_index = entry_index(l, b, n*l.w*l.h + j*l.w + i, 4 + 1);
int obj_index = entry_index(l, b, n*l.w*l.h + j*l.w + i, 4);
float objectness = l.output[obj_index];
int pred_class_id = get_yolo_class(l.output, l.classes, class_index, l.w*l.h, objectness);
if (class_id == pred_class_id) class_id_match = 1;
else class_id_match = 0;
float iou = box_iou(pred, truth); float iou = box_iou(pred, truth);
if (iou > best_iou) { //if (iou > best_iou) {
if (iou > best_iou && class_id_match == 1) {
best_iou = iou; best_iou = iou;
best_t = t; best_t = t;
} }

Write Preview
Loading…
Cancel
Save