| | |
| | | int detection_out_height(detection_layer layer) |
| | | #include "detection_layer.h" |
| | | #include "activations.h" |
| | | #include "softmax_layer.h" |
| | | #include "blas.h" |
| | | #include "cuda.h" |
| | | #include "utils.h" |
| | | #include <stdio.h> |
| | | #include <string.h> |
| | | #include <stdlib.h> |
| | | |
| | | int get_detection_layer_locations(detection_layer l) |
| | | { |
| | | return layer.size + layer.h*layer.stride; |
| | | return l.inputs / (l.classes+l.coords+l.joint+(l.background || l.objectness)); |
| | | } |
| | | |
| | | int detection_out_width(detection_layer layer) |
| | | int get_detection_layer_output_size(detection_layer l) |
| | | { |
| | | return layer.size + layer.w*layer.stride; |
| | | return get_detection_layer_locations(l)*((l.background || l.objectness) + l.classes + l.coords); |
| | | } |
| | | |
| | | detection_layer *make_detection_layer(int batch, int h, int w, int c, int n, int size, int stride, ACTIVATION activation) |
| | | detection_layer make_detection_layer(int batch, int inputs, int classes, int coords, int joint, int rescore, int background, int objectness) |
| | | { |
| | | int i; |
| | | size = 2*(size/2)+1; //HA! And you thought you'd use an even sized filter... |
| | | detection_layer *layer = calloc(1, sizeof(detection_layer)); |
| | | layer->h = h; |
| | | layer->w = w; |
| | | layer->c = c; |
| | | layer->n = n; |
| | | layer->batch = batch; |
| | | layer->stride = stride; |
| | | layer->size = size; |
| | | assert(c%n == 0); |
| | | detection_layer l = {0}; |
| | | l.type = DETECTION; |
| | | |
| | | l.batch = batch; |
| | | l.inputs = inputs; |
| | | l.classes = classes; |
| | | l.coords = coords; |
| | | l.rescore = rescore; |
| | | l.objectness = objectness; |
| | | l.joint = joint; |
| | | l.cost = calloc(1, sizeof(float)); |
| | | l.does_cost=1; |
| | | l.background = background; |
| | | int outputs = get_detection_layer_output_size(l); |
| | | l.outputs = outputs; |
| | | l.output = calloc(batch*outputs, sizeof(float)); |
| | | l.delta = calloc(batch*outputs, sizeof(float)); |
| | | #ifdef GPU |
| | | l.output_gpu = cuda_make_array(0, batch*outputs); |
| | | l.delta_gpu = cuda_make_array(0, batch*outputs); |
| | | #endif |
| | | |
| | | layer->filters = calloc(c*size*size, sizeof(float)); |
| | | layer->filter_updates = calloc(c*size*size, sizeof(float)); |
| | | layer->filter_momentum = calloc(c*size*size, sizeof(float)); |
| | | |
| | | float scale = 1./(size*size*c); |
| | | for(i = 0; i < c*n*size*size; ++i) layer->filters[i] = scale*(rand_uniform()); |
| | | |
| | | int out_h = detection_out_height(*layer); |
| | | int out_w = detection_out_width(*layer); |
| | | |
| | | layer->output = calloc(layer->batch * out_h * out_w * n, sizeof(float)); |
| | | layer->delta = calloc(layer->batch * out_h * out_w * n, sizeof(float)); |
| | | |
| | | layer->activation = activation; |
| | | |
| | | fprintf(stderr, "Convolutional Layer: %d x %d x %d image, %d filters -> %d x %d x %d image\n", h,w,c,n, out_h, out_w, n); |
| | | fprintf(stderr, "Detection Layer\n"); |
| | | srand(0); |
| | | |
| | | return layer; |
| | | return l; |
| | | } |
| | | |
| | | void forward_detection_layer(const detection_layer layer, float *in) |
| | | typedef struct{ |
| | | float dx, dy, dw, dh; |
| | | } dbox; |
| | | |
| | | dbox derivative(box a, box b) |
| | | { |
| | | int out_h = detection_out_height(layer); |
| | | int out_w = detection_out_width(layer); |
| | | int i,j,fh, fw,c; |
| | | memset(layer.output, 0, layer->batch*layer->n*out_h*out_w*sizeof(float)); |
| | | for(c = 0; c < layer.c; ++c){ |
| | | for(i = 0; i < layer.h; ++i){ |
| | | for(j = 0; j < layer.w; ++j){ |
| | | float val = layer->input[j+(i + c*layer.h)*layer.w]; |
| | | for(fh = 0; fh < layer.size; ++fh){ |
| | | for(fw = 0; fw < layer.size; ++fw){ |
| | | int h = i*layer.stride + fh; |
| | | int w = j*layer.stride + fw; |
| | | layer.output[w+(h+c/n*out_h)*out_w] += val*layer->filters[fw+(fh+c*layer.size)*layer.size]; |
| | | } |
| | | dbox d; |
| | | d.dx = 0; |
| | | d.dw = 0; |
| | | float l1 = a.x - a.w/2; |
| | | float l2 = b.x - b.w/2; |
| | | if (l1 > l2){ |
| | | d.dx -= 1; |
| | | d.dw += .5; |
| | | } |
| | | float r1 = a.x + a.w/2; |
| | | float r2 = b.x + b.w/2; |
| | | if(r1 < r2){ |
| | | d.dx += 1; |
| | | d.dw += .5; |
| | | } |
| | | if (l1 > r2) { |
| | | d.dx = -1; |
| | | d.dw = 0; |
| | | } |
| | | if (r1 < l2){ |
| | | d.dx = 1; |
| | | d.dw = 0; |
| | | } |
| | | |
| | | d.dy = 0; |
| | | d.dh = 0; |
| | | float t1 = a.y - a.h/2; |
| | | float t2 = b.y - b.h/2; |
| | | if (t1 > t2){ |
| | | d.dy -= 1; |
| | | d.dh += .5; |
| | | } |
| | | float b1 = a.y + a.h/2; |
| | | float b2 = b.y + b.h/2; |
| | | if(b1 < b2){ |
| | | d.dy += 1; |
| | | d.dh += .5; |
| | | } |
| | | if (t1 > b2) { |
| | | d.dy = -1; |
| | | d.dh = 0; |
| | | } |
| | | if (b1 < t2){ |
| | | d.dy = 1; |
| | | d.dh = 0; |
| | | } |
| | | return d; |
| | | } |
| | | |
| | | float overlap(float x1, float w1, float x2, float w2) |
| | | { |
| | | float l1 = x1 - w1/2; |
| | | float l2 = x2 - w2/2; |
| | | float left = l1 > l2 ? l1 : l2; |
| | | float r1 = x1 + w1/2; |
| | | float r2 = x2 + w2/2; |
| | | float right = r1 < r2 ? r1 : r2; |
| | | return right - left; |
| | | } |
| | | |
| | | float box_intersection(box a, box b) |
| | | { |
| | | float w = overlap(a.x, a.w, b.x, b.w); |
| | | float h = overlap(a.y, a.h, b.y, b.h); |
| | | if(w < 0 || h < 0) return 0; |
| | | float area = w*h; |
| | | return area; |
| | | } |
| | | |
| | | float box_union(box a, box b) |
| | | { |
| | | float i = box_intersection(a, b); |
| | | float u = a.w*a.h + b.w*b.h - i; |
| | | return u; |
| | | } |
| | | |
| | | float box_iou(box a, box b) |
| | | { |
| | | return box_intersection(a, b)/box_union(a, b); |
| | | } |
| | | |
| | | dbox dintersect(box a, box b) |
| | | { |
| | | float w = overlap(a.x, a.w, b.x, b.w); |
| | | float h = overlap(a.y, a.h, b.y, b.h); |
| | | dbox dover = derivative(a, b); |
| | | dbox di; |
| | | |
| | | di.dw = dover.dw*h; |
| | | di.dx = dover.dx*h; |
| | | di.dh = dover.dh*w; |
| | | di.dy = dover.dy*w; |
| | | |
| | | return di; |
| | | } |
| | | |
| | | dbox dunion(box a, box b) |
| | | { |
| | | dbox du; |
| | | |
| | | dbox di = dintersect(a, b); |
| | | du.dw = a.h - di.dw; |
| | | du.dh = a.w - di.dh; |
| | | du.dx = -di.dx; |
| | | du.dy = -di.dy; |
| | | |
| | | return du; |
| | | } |
| | | |
| | | dbox diou(box a, box b); |
| | | |
| | | void test_dunion() |
| | | { |
| | | box a = {0, 0, 1, 1}; |
| | | box dxa= {0+.0001, 0, 1, 1}; |
| | | box dya= {0, 0+.0001, 1, 1}; |
| | | box dwa= {0, 0, 1+.0001, 1}; |
| | | box dha= {0, 0, 1, 1+.0001}; |
| | | |
| | | box b = {.5, .5, .2, .2}; |
| | | dbox di = dunion(a,b); |
| | | printf("Union: %f %f %f %f\n", di.dx, di.dy, di.dw, di.dh); |
| | | float inter = box_union(a, b); |
| | | float xinter = box_union(dxa, b); |
| | | float yinter = box_union(dya, b); |
| | | float winter = box_union(dwa, b); |
| | | float hinter = box_union(dha, b); |
| | | xinter = (xinter - inter)/(.0001); |
| | | yinter = (yinter - inter)/(.0001); |
| | | winter = (winter - inter)/(.0001); |
| | | hinter = (hinter - inter)/(.0001); |
| | | printf("Union Manual %f %f %f %f\n", xinter, yinter, winter, hinter); |
| | | } |
| | | void test_dintersect() |
| | | { |
| | | box a = {0, 0, 1, 1}; |
| | | box dxa= {0+.0001, 0, 1, 1}; |
| | | box dya= {0, 0+.0001, 1, 1}; |
| | | box dwa= {0, 0, 1+.0001, 1}; |
| | | box dha= {0, 0, 1, 1+.0001}; |
| | | |
| | | box b = {.5, .5, .2, .2}; |
| | | dbox di = dintersect(a,b); |
| | | printf("Inter: %f %f %f %f\n", di.dx, di.dy, di.dw, di.dh); |
| | | float inter = box_intersection(a, b); |
| | | float xinter = box_intersection(dxa, b); |
| | | float yinter = box_intersection(dya, b); |
| | | float winter = box_intersection(dwa, b); |
| | | float hinter = box_intersection(dha, b); |
| | | xinter = (xinter - inter)/(.0001); |
| | | yinter = (yinter - inter)/(.0001); |
| | | winter = (winter - inter)/(.0001); |
| | | hinter = (hinter - inter)/(.0001); |
| | | printf("Inter Manual %f %f %f %f\n", xinter, yinter, winter, hinter); |
| | | } |
| | | |
| | | void test_box() |
| | | { |
| | | test_dintersect(); |
| | | test_dunion(); |
| | | box a = {0, 0, 1, 1}; |
| | | box dxa= {0+.00001, 0, 1, 1}; |
| | | box dya= {0, 0+.00001, 1, 1}; |
| | | box dwa= {0, 0, 1+.00001, 1}; |
| | | box dha= {0, 0, 1, 1+.00001}; |
| | | |
| | | box b = {.5, 0, .2, .2}; |
| | | |
| | | float iou = box_iou(a,b); |
| | | iou = (1-iou)*(1-iou); |
| | | printf("%f\n", iou); |
| | | dbox d = diou(a, b); |
| | | printf("%f %f %f %f\n", d.dx, d.dy, d.dw, d.dh); |
| | | |
| | | float xiou = box_iou(dxa, b); |
| | | float yiou = box_iou(dya, b); |
| | | float wiou = box_iou(dwa, b); |
| | | float hiou = box_iou(dha, b); |
| | | xiou = ((1-xiou)*(1-xiou) - iou)/(.00001); |
| | | yiou = ((1-yiou)*(1-yiou) - iou)/(.00001); |
| | | wiou = ((1-wiou)*(1-wiou) - iou)/(.00001); |
| | | hiou = ((1-hiou)*(1-hiou) - iou)/(.00001); |
| | | printf("manual %f %f %f %f\n", xiou, yiou, wiou, hiou); |
| | | } |
| | | |
| | | dbox diou(box a, box b) |
| | | { |
| | | float u = box_union(a,b); |
| | | float i = box_intersection(a,b); |
| | | dbox di = dintersect(a,b); |
| | | dbox du = dunion(a,b); |
| | | dbox dd = {0,0,0,0}; |
| | | |
| | | if(i <= 0 || 1) { |
| | | dd.dx = b.x - a.x; |
| | | dd.dy = b.y - a.y; |
| | | dd.dw = b.w - a.w; |
| | | dd.dh = b.h - a.h; |
| | | return dd; |
| | | } |
| | | |
| | | dd.dx = 2*pow((1-(i/u)),1)*(di.dx*u - du.dx*i)/(u*u); |
| | | dd.dy = 2*pow((1-(i/u)),1)*(di.dy*u - du.dy*i)/(u*u); |
| | | dd.dw = 2*pow((1-(i/u)),1)*(di.dw*u - du.dw*i)/(u*u); |
| | | dd.dh = 2*pow((1-(i/u)),1)*(di.dh*u - du.dh*i)/(u*u); |
| | | return dd; |
| | | } |
| | | |
| | | void forward_detection_layer(const detection_layer l, network_state state) |
| | | { |
| | | int in_i = 0; |
| | | int out_i = 0; |
| | | int locations = get_detection_layer_locations(l); |
| | | int i,j; |
| | | for(i = 0; i < l.batch*locations; ++i){ |
| | | int mask = (!state.truth || state.truth[out_i + (l.background || l.objectness) + l.classes + 2]); |
| | | float scale = 1; |
| | | if(l.joint) scale = state.input[in_i++]; |
| | | else if(l.objectness){ |
| | | l.output[out_i++] = 1-state.input[in_i++]; |
| | | scale = mask; |
| | | } |
| | | else if(l.background) l.output[out_i++] = scale*state.input[in_i++]; |
| | | |
| | | for(j = 0; j < l.classes; ++j){ |
| | | l.output[out_i++] = scale*state.input[in_i++]; |
| | | } |
| | | if(l.objectness){ |
| | | |
| | | }else if(l.background){ |
| | | softmax_array(l.output + out_i - l.classes-l.background, l.classes+l.background, l.output + out_i - l.classes-l.background); |
| | | activate_array(state.input+in_i, l.coords, LOGISTIC); |
| | | } |
| | | for(j = 0; j < l.coords; ++j){ |
| | | l.output[out_i++] = mask*state.input[in_i++]; |
| | | } |
| | | } |
| | | float avg_iou = 0; |
| | | int count = 0; |
| | | if(l.does_cost && state.train){ |
| | | *(l.cost) = 0; |
| | | int size = get_detection_layer_output_size(l) * l.batch; |
| | | memset(l.delta, 0, size * sizeof(float)); |
| | | for (i = 0; i < l.batch*locations; ++i) { |
| | | int classes = l.objectness+l.classes; |
| | | int offset = i*(classes+l.coords); |
| | | for (j = offset; j < offset+classes; ++j) { |
| | | *(l.cost) += pow(state.truth[j] - l.output[j], 2); |
| | | l.delta[j] = state.truth[j] - l.output[j]; |
| | | } |
| | | |
| | | box truth; |
| | | truth.x = state.truth[j+0]/7; |
| | | truth.y = state.truth[j+1]/7; |
| | | truth.w = pow(state.truth[j+2], 2); |
| | | truth.h = pow(state.truth[j+3], 2); |
| | | box out; |
| | | out.x = l.output[j+0]/7; |
| | | out.y = l.output[j+1]/7; |
| | | out.w = pow(l.output[j+2], 2); |
| | | out.h = pow(l.output[j+3], 2); |
| | | |
| | | if(!(truth.w*truth.h)) continue; |
| | | float iou = box_iou(out, truth); |
| | | avg_iou += iou; |
| | | ++count; |
| | | dbox delta = diou(out, truth); |
| | | |
| | | l.delta[j+0] = 10 * delta.dx/7; |
| | | l.delta[j+1] = 10 * delta.dy/7; |
| | | l.delta[j+2] = 10 * delta.dw * 2 * sqrt(out.w); |
| | | l.delta[j+3] = 10 * delta.dh * 2 * sqrt(out.h); |
| | | |
| | | |
| | | *(l.cost) += pow((1-iou), 2); |
| | | l.delta[j+0] = 4 * (state.truth[j+0] - l.output[j+0]); |
| | | l.delta[j+1] = 4 * (state.truth[j+1] - l.output[j+1]); |
| | | l.delta[j+2] = 4 * (state.truth[j+2] - l.output[j+2]); |
| | | l.delta[j+3] = 4 * (state.truth[j+3] - l.output[j+3]); |
| | | if(l.rescore){ |
| | | for (j = offset; j < offset+classes; ++j) { |
| | | if(state.truth[j]) state.truth[j] = iou; |
| | | l.delta[j] = state.truth[j] - l.output[j]; |
| | | } |
| | | } |
| | | } |
| | | printf("Avg IOU: %f\n", avg_iou/count); |
| | | } |
| | | } |
| | | |
| | | void backward_detection_layer(const detection_layer layer, float *delta) |
| | | void backward_detection_layer(const detection_layer l, network_state state) |
| | | { |
| | | int locations = get_detection_layer_locations(l); |
| | | int i,j; |
| | | int in_i = 0; |
| | | int out_i = 0; |
| | | for(i = 0; i < l.batch*locations; ++i){ |
| | | float scale = 1; |
| | | float latent_delta = 0; |
| | | if(l.joint) scale = state.input[in_i++]; |
| | | else if (l.objectness) state.delta[in_i++] = -l.delta[out_i++]; |
| | | else if (l.background) state.delta[in_i++] = scale*l.delta[out_i++]; |
| | | for(j = 0; j < l.classes; ++j){ |
| | | latent_delta += state.input[in_i]*l.delta[out_i]; |
| | | state.delta[in_i++] = scale*l.delta[out_i++]; |
| | | } |
| | | |
| | | if (l.objectness) { |
| | | |
| | | }else if (l.background) gradient_array(l.output + out_i, l.coords, LOGISTIC, l.delta + out_i); |
| | | for(j = 0; j < l.coords; ++j){ |
| | | state.delta[in_i++] = l.delta[out_i++]; |
| | | } |
| | | if(l.joint) state.delta[in_i-l.coords-l.classes-l.joint] = latent_delta; |
| | | } |
| | | } |
| | | |
| | | #ifdef GPU |
| | | |
| | | void forward_detection_layer_gpu(const detection_layer l, network_state state) |
| | | { |
| | | int outputs = get_detection_layer_output_size(l); |
| | | float *in_cpu = calloc(l.batch*l.inputs, sizeof(float)); |
| | | float *truth_cpu = 0; |
| | | if(state.truth){ |
| | | truth_cpu = calloc(l.batch*outputs, sizeof(float)); |
| | | cuda_pull_array(state.truth, truth_cpu, l.batch*outputs); |
| | | } |
| | | cuda_pull_array(state.input, in_cpu, l.batch*l.inputs); |
| | | network_state cpu_state; |
| | | cpu_state.train = state.train; |
| | | cpu_state.truth = truth_cpu; |
| | | cpu_state.input = in_cpu; |
| | | forward_detection_layer(l, cpu_state); |
| | | cuda_push_array(l.output_gpu, l.output, l.batch*outputs); |
| | | cuda_push_array(l.delta_gpu, l.delta, l.batch*outputs); |
| | | free(cpu_state.input); |
| | | if(cpu_state.truth) free(cpu_state.truth); |
| | | } |
| | | |
| | | void backward_detection_layer_gpu(detection_layer l, network_state state) |
| | | { |
| | | int outputs = get_detection_layer_output_size(l); |
| | | |
| | | float *in_cpu = calloc(l.batch*l.inputs, sizeof(float)); |
| | | float *delta_cpu = calloc(l.batch*l.inputs, sizeof(float)); |
| | | float *truth_cpu = 0; |
| | | if(state.truth){ |
| | | truth_cpu = calloc(l.batch*outputs, sizeof(float)); |
| | | cuda_pull_array(state.truth, truth_cpu, l.batch*outputs); |
| | | } |
| | | network_state cpu_state; |
| | | cpu_state.train = state.train; |
| | | cpu_state.input = in_cpu; |
| | | cpu_state.truth = truth_cpu; |
| | | cpu_state.delta = delta_cpu; |
| | | |
| | | cuda_pull_array(state.input, in_cpu, l.batch*l.inputs); |
| | | cuda_pull_array(l.delta_gpu, l.delta, l.batch*outputs); |
| | | backward_detection_layer(l, cpu_state); |
| | | cuda_push_array(state.delta, delta_cpu, l.batch*l.inputs); |
| | | |
| | | free(in_cpu); |
| | | free(delta_cpu); |
| | | } |
| | | #endif |
| | | |