From aebe937710ced03d03f73ab23f410f29685655c1 Mon Sep 17 00:00:00 2001
From: Joseph Redmon <pjreddie@gmail.com>
Date: Thu, 11 Aug 2016 18:54:24 +0000
Subject: [PATCH] what do you even write here?
---
src/region_layer.c | 332 +++++++++++++++++++++++++++++++++++-------------------
1 files changed, 213 insertions(+), 119 deletions(-)
diff --git a/src/region_layer.c b/src/region_layer.c
index ecb89c6..5fe37c5 100644
--- a/src/region_layer.c
+++ b/src/region_layer.c
@@ -10,24 +10,23 @@
#include <string.h>
#include <stdlib.h>
-region_layer make_region_layer(int batch, int inputs, int n, int side, int classes, int coords, int rescore)
+region_layer make_region_layer(int batch, int w, int h, int n, int classes, int coords)
{
region_layer l = {0};
l.type = REGION;
l.n = n;
l.batch = batch;
- l.inputs = inputs;
+ l.h = h;
+ l.w = w;
l.classes = classes;
l.coords = coords;
- l.rescore = rescore;
- l.side = side;
- assert(side*side*((1 + l.coords)*l.n + l.classes) == inputs);
l.cost = calloc(1, sizeof(float));
- l.outputs = l.inputs;
- l.truths = l.side*l.side*(1+l.coords+l.classes);
- l.output = calloc(batch*l.outputs, sizeof(float));
+ l.outputs = h*w*n*(classes + coords + 1);
+ l.inputs = l.outputs;
+ l.truths = 30*(5);
l.delta = calloc(batch*l.outputs, sizeof(float));
+ l.output = calloc(batch*l.outputs, sizeof(float));
#ifdef GPU
l.output_gpu = cuda_make_array(l.output, batch*l.outputs);
l.delta_gpu = cuda_make_array(l.delta, batch*l.outputs);
@@ -39,120 +38,208 @@
return l;
}
+box get_region_box2(float *x, int index, int i, int j, int w, int h)
+{
+ float aspect = exp(x[index+0]);
+ float scale = logistic_activate(x[index+1]);
+ float move_x = x[index+2];
+ float move_y = x[index+3];
+
+ box b;
+ b.w = sqrt(scale * aspect);
+ b.h = b.w * 1./aspect;
+ b.x = move_x * b.w + (i + .5)/w;
+ b.y = move_y * b.h + (j + .5)/h;
+ return b;
+}
+
+float delta_region_box2(box truth, float *output, int index, int i, int j, int w, int h, float *delta)
+{
+ box pred = get_region_box2(output, index, i, j, w, h);
+ float iou = box_iou(pred, truth);
+ float true_aspect = truth.w/truth.h;
+ float true_scale = truth.w*truth.h;
+
+ float true_dx = (truth.x - (i+.5)/w) / truth.w;
+ float true_dy = (truth.y - (j+.5)/h) / truth.h;
+ delta[index + 0] = (true_aspect - exp(output[index + 0])) * exp(output[index + 0]);
+ delta[index + 1] = (true_scale - logistic_activate(output[index + 1])) * logistic_gradient(logistic_activate(output[index + 1]));
+ delta[index + 2] = true_dx - output[index + 2];
+ delta[index + 3] = true_dy - output[index + 3];
+ return iou;
+}
+
+box get_region_box(float *x, int index, int i, int j, int w, int h, int adjust, int logistic)
+{
+ box b;
+ b.x = (x[index + 0] + i + .5)/w;
+ b.y = (x[index + 1] + j + .5)/h;
+ b.w = x[index + 2];
+ b.h = x[index + 3];
+ if(logistic){
+ b.w = logistic_activate(x[index + 2]);
+ b.h = logistic_activate(x[index + 3]);
+ }
+ if(adjust && b.w < .01) b.w = .01;
+ if(adjust && b.h < .01) b.h = .01;
+ return b;
+}
+
+float delta_region_box(box truth, float *output, int index, int i, int j, int w, int h, float *delta, int logistic, float scale)
+{
+ box pred = get_region_box(output, index, i, j, w, h, 0, logistic);
+ float iou = box_iou(pred, truth);
+
+ delta[index + 0] = scale * (truth.x - pred.x);
+ delta[index + 1] = scale * (truth.y - pred.y);
+ delta[index + 2] = scale * ((truth.w - pred.w)*(logistic ? logistic_gradient(pred.w) : 1));
+ delta[index + 3] = scale * ((truth.h - pred.h)*(logistic ? logistic_gradient(pred.h) : 1));
+ return iou;
+}
+
+float logit(float x)
+{
+ return log(x/(1.-x));
+}
+
+float tisnan(float x)
+{
+ return (x != x);
+}
+
+#define LOG 1
+
void forward_region_layer(const region_layer l, network_state state)
{
- int locations = l.side*l.side;
- int i,j;
+ int i,j,b,t,n;
+ int size = l.coords + l.classes + 1;
memcpy(l.output, state.input, l.outputs*l.batch*sizeof(float));
- int b;
- if (l.softmax){
- for(b = 0; b < l.batch; ++b){
- int index = b*l.inputs;
- for (i = 0; i < locations; ++i) {
- int offset = i*l.classes;
- softmax_array(l.output + index + offset, l.classes,
- l.output + index + offset);
- }
- int offset = locations*l.classes;
- activate_array(l.output + index + offset, locations*l.n*(1+l.coords), LOGISTIC);
- }
- }
-
- if(state.train){
- float avg_iou = 0;
- float avg_cat = 0;
- float avg_obj = 0;
- float avg_anyobj = 0;
- int count = 0;
- *(l.cost) = 0;
- int size = l.inputs * l.batch;
- memset(l.delta, 0, size * sizeof(float));
- for (b = 0; b < l.batch; ++b){
- int index = b*l.inputs;
- for (i = 0; i < locations; ++i) {
- int truth_index = (b*locations + i)*(1+l.coords+l.classes);
- int is_obj = state.truth[truth_index];
- for (j = 0; j < l.n; ++j) {
- int p_index = index + locations*l.classes + i*l.n + j;
- l.delta[p_index] = l.noobject_scale*(0 - l.output[p_index]);
- *(l.cost) += l.noobject_scale*pow(l.output[p_index], 2);
- avg_anyobj += l.output[p_index];
- }
-
- int best_index = -1;
- float best_iou = 0;
- float best_rmse = 4;
-
- if (!is_obj) continue;
-
- int class_index = index + i*l.classes;
- for(j = 0; j < l.classes; ++j) {
- l.delta[class_index+j] = l.class_scale * (state.truth[truth_index+1+j] - l.output[class_index+j]);
- *(l.cost) += l.class_scale * pow(state.truth[truth_index+1+j] - l.output[class_index+j], 2);
- if(state.truth[truth_index + 1 + j]) avg_cat += l.output[class_index+j];
- }
-
- box truth = float_to_box(state.truth + truth_index + 1 + l.classes);
- truth.x /= l.side;
- truth.y /= l.side;
-
- for(j = 0; j < l.n; ++j){
- int box_index = index + locations*(l.classes + l.n) + (i*l.n + j) * l.coords;
- box out = float_to_box(l.output + box_index);
- out.x /= l.side;
- out.y /= l.side;
-
- if (l.sqrt){
- out.w = out.w*out.w;
- out.h = out.h*out.h;
- }
-
- float iou = box_iou(out, truth);
- float rmse = box_rmse(out, truth);
- if(best_iou > 0 || iou > 0){
- if(iou > best_iou){
- best_iou = iou;
- best_index = j;
- }
- }else{
- if(rmse < best_rmse){
- best_rmse = rmse;
- best_index = j;
- }
- }
- }
- int p_index = index + locations*l.classes + i*l.n + best_index;
- *(l.cost) -= l.noobject_scale * pow(l.output[p_index], 2);
- *(l.cost) += l.object_scale * pow(1-l.output[p_index], 2);
- avg_obj += l.output[p_index];
- l.delta[p_index+0] = l.object_scale * (1.-l.output[p_index]);
-
- if(l.rescore){
- l.delta[p_index+0] = l.object_scale * (best_iou - l.output[p_index]);
- }
-
- int box_index = index + locations*(l.classes + l.n) + (i*l.n + best_index) * l.coords;
- int tbox_index = truth_index + 1 + l.classes;
- l.delta[box_index+0] = l.coord_scale*(state.truth[tbox_index + 0] - l.output[box_index + 0]);
- l.delta[box_index+1] = l.coord_scale*(state.truth[tbox_index + 1] - l.output[box_index + 1]);
- l.delta[box_index+2] = l.coord_scale*(state.truth[tbox_index + 2] - l.output[box_index + 2]);
- l.delta[box_index+3] = l.coord_scale*(state.truth[tbox_index + 3] - l.output[box_index + 3]);
- if(l.sqrt){
- l.delta[box_index+2] = l.coord_scale*(sqrt(state.truth[tbox_index + 2]) - l.output[box_index + 2]);
- l.delta[box_index+3] = l.coord_scale*(sqrt(state.truth[tbox_index + 3]) - l.output[box_index + 3]);
- }
-
- *(l.cost) += pow(1-best_iou, 2);
- avg_iou += best_iou;
- ++count;
- }
+ reorg(l.output, l.w*l.h, size*l.n, l.batch, 1);
+ for (b = 0; b < l.batch; ++b){
+ for(i = 0; i < l.h*l.w*l.n; ++i){
+ int index = size*i + b*l.outputs;
+ l.output[index + 4] = logistic_activate(l.output[index + 4]);
if(l.softmax){
- gradient_array(l.output + index + locations*l.classes, locations*l.n*(1+l.coords),
- LOGISTIC, l.delta + index + locations*l.classes);
+ softmax_array(l.output + index + 5, l.classes, 1, l.output + index + 5);
}
}
- printf("Region Avg IOU: %f, Avg Cat Pred: %f, Avg Obj: %f, Avg Any: %f, count: %d\n", avg_iou/count, avg_cat/count, avg_obj/count, avg_anyobj/(l.batch*locations*l.n), count);
}
+ if(!state.train) return;
+ memset(l.delta, 0, l.outputs * l.batch * sizeof(float));
+ float avg_iou = 0;
+ float avg_cat = 0;
+ float avg_obj = 0;
+ float avg_anyobj = 0;
+ int count = 0;
+ *(l.cost) = 0;
+ for (b = 0; b < l.batch; ++b) {
+ for (j = 0; j < l.h; ++j) {
+ for (i = 0; i < l.w; ++i) {
+ for (n = 0; n < l.n; ++n) {
+ int index = size*(j*l.w*l.n + i*l.n + n) + b*l.outputs;
+ box pred = get_region_box(l.output, index, i, j, l.w, l.h, 1, LOG);
+ float best_iou = 0;
+ for(t = 0; t < 30; ++t){
+ box truth = float_to_box(state.truth + t*5 + b*l.truths);
+ if(!truth.x) break;
+ float iou = box_iou(pred, truth);
+ if (iou > best_iou) best_iou = iou;
+ }
+ avg_anyobj += l.output[index + 4];
+ l.delta[index + 4] = l.noobject_scale * ((0 - l.output[index + 4]) * logistic_gradient(l.output[index + 4]));
+ if(best_iou > .5) l.delta[index + 4] = 0;
+
+ if(*(state.net.seen) < 6400){
+ box truth = {0};
+ truth.x = (i + .5)/l.w;
+ truth.y = (j + .5)/l.h;
+ truth.w = .5;
+ truth.h = .5;
+ delta_region_box(truth, l.output, index, i, j, l.w, l.h, l.delta, LOG, 1);
+ }
+ }
+ }
+ }
+ for(t = 0; t < 30; ++t){
+ box truth = float_to_box(state.truth + t*5 + b*l.truths);
+ int class = state.truth[t*5 + b*l.truths + 4];
+ if(!truth.x) break;
+ float best_iou = 0;
+ int best_index = 0;
+ int best_n = 0;
+ i = (truth.x * l.w);
+ j = (truth.y * l.h);
+ //printf("%d %f %d %f\n", i, truth.x*l.w, j, truth.y*l.h);
+ box truth_shift = truth;
+ truth_shift.x = 0;
+ truth_shift.y = 0;
+ printf("index %d %d\n",i, j);
+ for(n = 0; n < l.n; ++n){
+ int index = size*(j*l.w*l.n + i*l.n + n) + b*l.outputs;
+ box pred = get_region_box(l.output, index, i, j, l.w, l.h, 1, LOG);
+ printf("pred: (%f, %f) %f x %f\n", pred.x, pred.y, pred.w, pred.h);
+ pred.x = 0;
+ pred.y = 0;
+ float iou = box_iou(pred, truth_shift);
+ if (iou > best_iou){
+ best_index = index;
+ best_iou = iou;
+ best_n = n;
+ }
+ }
+ printf("%d %f (%f, %f) %f x %f\n", best_n, best_iou, truth.x, truth.y, truth.w, truth.h);
+
+ float iou = delta_region_box(truth, l.output, best_index, i, j, l.w, l.h, l.delta, LOG, l.coord_scale);
+ avg_iou += iou;
+
+ //l.delta[best_index + 4] = iou - l.output[best_index + 4];
+ avg_obj += l.output[best_index + 4];
+ l.delta[best_index + 4] = l.object_scale * (1 - l.output[best_index + 4]) * logistic_gradient(l.output[best_index + 4]);
+ if (l.rescore) {
+ l.delta[best_index + 4] = l.object_scale * (iou - l.output[best_index + 4]) * logistic_gradient(l.output[best_index + 4]);
+ }
+ //printf("%f\n", l.delta[best_index+1]);
+ /*
+ if(isnan(l.delta[best_index+1])){
+ printf("%f\n", true_scale);
+ printf("%f\n", l.output[best_index + 1]);
+ printf("%f\n", truth.w);
+ printf("%f\n", truth.h);
+ error("bad");
+ }
+ */
+ for(n = 0; n < l.classes; ++n){
+ l.delta[best_index + 5 + n] = l.class_scale * (((n == class)?1 : 0) - l.output[best_index + 5 + n]);
+ if(n == class) avg_cat += l.output[best_index + 5 + n];
+ }
+ /*
+ if(0){
+ printf("truth: %f %f %f %f\n", truth.x, truth.y, truth.w, truth.h);
+ printf("pred: %f %f %f %f\n\n", pred.x, pred.y, pred.w, pred.h);
+ float aspect = exp(true_aspect);
+ float scale = logistic_activate(true_scale);
+ float move_x = true_dx;
+ float move_y = true_dy;
+
+ box b;
+ b.w = sqrt(scale * aspect);
+ b.h = b.w * 1./aspect;
+ b.x = move_x * b.w + (i + .5)/l.w;
+ b.y = move_y * b.h + (j + .5)/l.h;
+ printf("%f %f\n", b.x, truth.x);
+ printf("%f %f\n", b.y, truth.y);
+ printf("%f %f\n", b.w, truth.w);
+ printf("%f %f\n", b.h, truth.h);
+ //printf("%f\n", box_iou(b, truth));
+ }
+ */
+ ++count;
+ }
+ }
+ printf("\n");
+ reorg(l.delta, l.w*l.h, size*l.n, l.batch, 0);
+ *(l.cost) = pow(mag_array(l.delta, l.outputs * l.batch), 2);
+ printf("Region Avg IOU: %f, Class: %f, Obj: %f, No Obj: %f, count: %d\n", avg_iou/count, avg_cat/count, avg_obj/count, avg_anyobj/(l.w*l.h*l.n*l.batch), count);
}
void backward_region_layer(const region_layer l, network_state state)
@@ -164,28 +251,35 @@
void forward_region_layer_gpu(const region_layer l, network_state state)
{
+ /*
+ if(!state.train){
+ copy_ongpu(l.batch*l.inputs, state.input, 1, l.output_gpu, 1);
+ return;
+ }
+ */
+
float *in_cpu = calloc(l.batch*l.inputs, sizeof(float));
float *truth_cpu = 0;
if(state.truth){
- int num_truth = l.batch*l.side*l.side*(1+l.coords+l.classes);
+ int num_truth = l.batch*l.truths;
truth_cpu = calloc(num_truth, sizeof(float));
cuda_pull_array(state.truth, truth_cpu, num_truth);
}
cuda_pull_array(state.input, in_cpu, l.batch*l.inputs);
- network_state cpu_state;
+ network_state cpu_state = state;
cpu_state.train = state.train;
cpu_state.truth = truth_cpu;
cpu_state.input = in_cpu;
forward_region_layer(l, cpu_state);
cuda_push_array(l.output_gpu, l.output, l.batch*l.outputs);
- cuda_push_array(l.delta_gpu, l.delta, l.batch*l.inputs);
+ cuda_push_array(l.delta_gpu, l.delta, l.batch*l.outputs);
free(cpu_state.input);
if(cpu_state.truth) free(cpu_state.truth);
}
void backward_region_layer_gpu(region_layer l, network_state state)
{
- axpy_ongpu(l.batch*l.inputs, 1, l.delta_gpu, 1, state.delta, 1);
+ axpy_ongpu(l.batch*l.outputs, 1, l.delta_gpu, 1, state.delta, 1);
//copy_ongpu(l.batch*l.inputs, l.delta_gpu, 1, state.delta, 1);
}
#endif
--
Gitblit v1.10.0