refactoring and added DARK ZONE

Joseph Redmon

2015-03-12 dcb000b553d051429a49c8729dc5b1af632e8532

 src/network.c

@@ -4,6 +4,7 @@
#include "image.h"
#include "data.h"
#include "utils.h"
#include "params.h"

#include "crop_layer.h"
#include "connected_layer.h"
@@ -13,7 +14,6 @@
#include "maxpool_layer.h"
#include "cost_layer.h"
#include "normalization_layer.h"
#include "freeweight_layer.h"
#include "softmax_layer.h"
#include "dropout_layer.h"

@@ -36,8 +36,6 @@
            return "normalization";
        case DROPOUT:
            return "dropout";
        case FREEWEIGHT:
            return "freeweight";
        case CROP:
            return "crop";
        case COST:
@@ -48,16 +46,18 @@
    return "none";
}

network make_network(int n, int batch)
network make_network(int n)
{
    network net;
    net.n = n;
    net.batch = batch;
    net.layers = calloc(net.n, sizeof(void *));
    net.types = calloc(net.n, sizeof(LAYER_TYPE));
    net.outputs = 0;
    net.output = 0;
    net.seen = 0;
    net.batch = 0;
    net.inputs = 0;
    net.h = net.w = net.c = 0;
    #ifdef GPU
    net.input_gpu = calloc(1, sizeof(float *));
    net.truth_gpu = calloc(1, sizeof(float *));
@@ -65,68 +65,41 @@
    return net;
}

void forward_network(network net, float *input, float *truth, int train)
void forward_network(network net, network_state state)
{
    int i;
    for(i = 0; i < net.n; ++i){
        if(net.types[i] == CONVOLUTIONAL){
            convolutional_layer layer = *(convolutional_layer *)net.layers[i];
            forward_convolutional_layer(layer, input);
            input = layer.output;
            forward_convolutional_layer(*(convolutional_layer *)net.layers[i], state);
        }
        else if(net.types[i] == DECONVOLUTIONAL){
            deconvolutional_layer layer = *(deconvolutional_layer *)net.layers[i];
            forward_deconvolutional_layer(layer, input);
            input = layer.output;
            forward_deconvolutional_layer(*(deconvolutional_layer *)net.layers[i], state);
        }
        else if(net.types[i] == DETECTION){
            detection_layer layer = *(detection_layer *)net.layers[i];
            forward_detection_layer(layer, input, truth);
            input = layer.output;
            forward_detection_layer(*(detection_layer *)net.layers[i], state);
        }
        else if(net.types[i] == CONNECTED){
            connected_layer layer = *(connected_layer *)net.layers[i];
            forward_connected_layer(layer, input);
            input = layer.output;
            forward_connected_layer(*(connected_layer *)net.layers[i], state);
        }
        else if(net.types[i] == CROP){
            crop_layer layer = *(crop_layer *)net.layers[i];
            forward_crop_layer(layer, train, input);
            input = layer.output;
            forward_crop_layer(*(crop_layer *)net.layers[i], state);
        }
        else if(net.types[i] == COST){
            cost_layer layer = *(cost_layer *)net.layers[i];
            forward_cost_layer(layer, input, truth);
            forward_cost_layer(*(cost_layer *)net.layers[i], state);
        }
        else if(net.types[i] == SOFTMAX){
            softmax_layer layer = *(softmax_layer *)net.layers[i];
            forward_softmax_layer(layer, input);
            input = layer.output;
            forward_softmax_layer(*(softmax_layer *)net.layers[i], state);
        }
        else if(net.types[i] == MAXPOOL){
            maxpool_layer layer = *(maxpool_layer *)net.layers[i];
            forward_maxpool_layer(layer, input);
            input = layer.output;
            forward_maxpool_layer(*(maxpool_layer *)net.layers[i], state);
        }
        else if(net.types[i] == NORMALIZATION){
            normalization_layer layer = *(normalization_layer *)net.layers[i];
            forward_normalization_layer(layer, input);
            input = layer.output;
            forward_normalization_layer(*(normalization_layer *)net.layers[i], state);
        }
        else if(net.types[i] == DROPOUT){
            if(!train) continue;
            dropout_layer layer = *(dropout_layer *)net.layers[i];
            forward_dropout_layer(layer, input);
            input = layer.output;
            forward_dropout_layer(*(dropout_layer *)net.layers[i], state);
        }
        else if(net.types[i] == FREEWEIGHT){
            if(!train) continue;
            //freeweight_layer layer = *(freeweight_layer *)net.layers[i];
            //forward_freeweight_layer(layer, input);
        }
        //char buff[256];
        //sprintf(buff, "layer %d", i);
        //cuda_compare(get_network_output_gpu_layer(net, i), input, get_network_output_size_layer(net, i)*net.batch, buff);
        state.input = get_network_output_layer(net, i);
    }
}

@@ -136,15 +109,15 @@
    for(i = 0; i < net.n; ++i){
        if(net.types[i] == CONVOLUTIONAL){
            convolutional_layer layer = *(convolutional_layer *)net.layers[i];
            update_convolutional_layer(layer);
            update_convolutional_layer(layer, net.learning_rate, net.momentum, net.decay);
        }
        else if(net.types[i] == DECONVOLUTIONAL){
            deconvolutional_layer layer = *(deconvolutional_layer *)net.layers[i];
            update_deconvolutional_layer(layer);
            update_deconvolutional_layer(layer, net.learning_rate, net.momentum, net.decay);
        }
        else if(net.types[i] == CONNECTED){
            connected_layer layer = *(connected_layer *)net.layers[i];
            update_connected_layer(layer);
            update_connected_layer(layer, net.learning_rate, net.momentum, net.decay);
        }
    }
}
@@ -152,37 +125,27 @@
float *get_network_output_layer(network net, int i)
{
    if(net.types[i] == CONVOLUTIONAL){
        convolutional_layer layer = *(convolutional_layer *)net.layers[i];
        return layer.output;
        return ((convolutional_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == DECONVOLUTIONAL){
        deconvolutional_layer layer = *(deconvolutional_layer *)net.layers[i];
        return layer.output;
        return ((deconvolutional_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == MAXPOOL){
        maxpool_layer layer = *(maxpool_layer *)net.layers[i];
        return layer.output;
        return ((maxpool_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == DETECTION){
        detection_layer layer = *(detection_layer *)net.layers[i];
        return layer.output;
        return ((detection_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == SOFTMAX){
        softmax_layer layer = *(softmax_layer *)net.layers[i];
        return layer.output;
        return ((softmax_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == DROPOUT){
        dropout_layer layer = *(dropout_layer *)net.layers[i];
        return layer.output;
    } else if(net.types[i] == FREEWEIGHT){
        return get_network_output_layer(net, i-1);
    } else if(net.types[i] == CONNECTED){
        connected_layer layer = *(connected_layer *)net.layers[i];
        return layer.output;
        return ((connected_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == CROP){
        crop_layer layer = *(crop_layer *)net.layers[i];
        return layer.output;
        return ((crop_layer *)net.layers[i]) -> output;
    } else if(net.types[i] == NORMALIZATION){
        normalization_layer layer = *(normalization_layer *)net.layers[i];
        return layer.output;
        return ((normalization_layer *)net.layers[i]) -> output;
    }
    return 0;
}

float *get_network_output(network net)
{
    int i;
@@ -210,8 +173,6 @@
    } else if(net.types[i] == DROPOUT){
        if(i == 0) return 0;
        return get_network_delta_layer(net, i-1);
    } else if(net.types[i] == FREEWEIGHT){
        return get_network_delta_layer(net, i-1);
    } else if(net.types[i] == CONNECTED){
        connected_layer layer = *(connected_layer *)net.layers[i];
        return layer.delta;
@@ -257,54 +218,53 @@
    return max_index(out, k);
}

void backward_network(network net, float *input, float *truth)
void backward_network(network net, network_state state)
{
    int i;
    float *prev_input;
    float *prev_delta;
    float *original_input = state.input;
    for(i = net.n-1; i >= 0; --i){
        if(i == 0){
            prev_input = input;
            prev_delta = 0;
            state.input = original_input;
            state.delta = 0;
        }else{
            prev_input = get_network_output_layer(net, i-1);
            prev_delta = get_network_delta_layer(net, i-1);
            state.input = get_network_output_layer(net, i-1);
            state.delta = get_network_delta_layer(net, i-1);
        }

        if(net.types[i] == CONVOLUTIONAL){
            convolutional_layer layer = *(convolutional_layer *)net.layers[i];
            backward_convolutional_layer(layer, prev_input, prev_delta);
            backward_convolutional_layer(layer, state);
        } else if(net.types[i] == DECONVOLUTIONAL){
            deconvolutional_layer layer = *(deconvolutional_layer *)net.layers[i];
            backward_deconvolutional_layer(layer, prev_input, prev_delta);
            backward_deconvolutional_layer(layer, state);
        }
        else if(net.types[i] == MAXPOOL){
            maxpool_layer layer = *(maxpool_layer *)net.layers[i];
            if(i != 0) backward_maxpool_layer(layer, prev_delta);
            if(i != 0) backward_maxpool_layer(layer, state);
        }
        else if(net.types[i] == DROPOUT){
            dropout_layer layer = *(dropout_layer *)net.layers[i];
            backward_dropout_layer(layer, prev_delta);
            backward_dropout_layer(layer, state);
        }
        else if(net.types[i] == DETECTION){
            detection_layer layer = *(detection_layer *)net.layers[i];
            backward_detection_layer(layer, prev_input, prev_delta);
            backward_detection_layer(layer, state);
        }
        else if(net.types[i] == NORMALIZATION){
            normalization_layer layer = *(normalization_layer *)net.layers[i];
            if(i != 0) backward_normalization_layer(layer, prev_input, prev_delta);
            if(i != 0) backward_normalization_layer(layer, state);
        }
        else if(net.types[i] == SOFTMAX){
            softmax_layer layer = *(softmax_layer *)net.layers[i];
            if(i != 0) backward_softmax_layer(layer, prev_delta);
            if(i != 0) backward_softmax_layer(layer, state);
        }
        else if(net.types[i] == CONNECTED){
            connected_layer layer = *(connected_layer *)net.layers[i];
            backward_connected_layer(layer, prev_input, prev_delta);
            backward_connected_layer(layer, state);
        }
        else if(net.types[i] == COST){
            cost_layer layer = *(cost_layer *)net.layers[i];
            backward_cost_layer(layer, prev_input, prev_delta);
            backward_cost_layer(layer, state);
        }
    }
}
@@ -314,8 +274,12 @@
    #ifdef GPU
    if(gpu_index >= 0) return train_network_datum_gpu(net, x, y);
    #endif
    forward_network(net, x, y, 1);
    backward_network(net, x, y);
    network_state state;
    state.input = x;
    state.truth = y;
    state.train = 1;
    forward_network(net, state);
    backward_network(net, state);
    float error = get_network_cost(net);
    update_network(net);
    return error;
@@ -361,15 +325,17 @@
float train_network_batch(network net, data d, int n)
{
    int i,j;
    network_state state;
    state.train = 1;
    float sum = 0;
    int batch = 2;
    for(i = 0; i < n; ++i){
        for(j = 0; j < batch; ++j){
            int index = rand()%d.X.rows;
            float *x = d.X.vals[index];
            float *y = d.y.vals[index];
            forward_network(net, x, y, 1);
            backward_network(net, x, y);
            state.input = d.X.vals[index];
            state.truth = d.y.vals[index];
            forward_network(net, state);
            backward_network(net, state);
            sum += get_network_cost(net);
        }
        update_network(net);
@@ -377,28 +343,6 @@
    return (float)sum/(n*batch);
}

void set_learning_network(network *net, float rate, float momentum, float decay)
{
    int i;
    net->learning_rate=rate;
    net->momentum = momentum;
    net->decay = decay;
    for(i = 0; i < net->n; ++i){
        if(net->types[i] == CONVOLUTIONAL){
            convolutional_layer *layer = (convolutional_layer *)net->layers[i];
            layer->learning_rate=rate;
            layer->momentum = momentum;
            layer->decay = decay;
        }
        else if(net->types[i] == CONNECTED){
            connected_layer *layer = (connected_layer *)net->layers[i];
            layer->learning_rate=rate;
            layer->momentum = momentum;
            layer->decay = decay;
        }
    }
}

void set_batch_network(network *net, int b)
{
    net->batch = b;
@@ -425,10 +369,6 @@
            detection_layer *layer = (detection_layer *) net->layers[i];
            layer->batch = b;
        }
        else if(net->types[i] == FREEWEIGHT){
            freeweight_layer *layer = (freeweight_layer *) net->layers[i];
            layer->batch = b;
        }
        else if(net->types[i] == SOFTMAX){
            softmax_layer *layer = (softmax_layer *)net->layers[i];
            layer->batch = b;
@@ -472,15 +412,11 @@
        crop_layer layer = *(crop_layer *) net.layers[i];
        return layer.c*layer.h*layer.w;
    }
    else if(net.types[i] == FREEWEIGHT){
        freeweight_layer layer = *(freeweight_layer *) net.layers[i];
        return layer.inputs;
    }
    else if(net.types[i] == SOFTMAX){
        softmax_layer layer = *(softmax_layer *)net.layers[i];
        return layer.inputs;
    }
    printf("Can't find input size\n");
    fprintf(stderr, "Can't find input size\n");
    return 0;
}

@@ -505,7 +441,7 @@
        image output = get_maxpool_image(layer);
        return output.h*output.w*output.c;
    }
     else if(net.types[i] == CROP){
    else if(net.types[i] == CROP){
        crop_layer layer = *(crop_layer *) net.layers[i];
        return layer.c*layer.crop_height*layer.crop_width;
    }
@@ -517,15 +453,11 @@
        dropout_layer layer = *(dropout_layer *) net.layers[i];
        return layer.inputs;
    }
    else if(net.types[i] == FREEWEIGHT){
        freeweight_layer layer = *(freeweight_layer *) net.layers[i];
        return layer.inputs;
    }
    else if(net.types[i] == SOFTMAX){
        softmax_layer layer = *(softmax_layer *)net.layers[i];
        return layer.inputs;
    }
    printf("Can't find output size\n");
    fprintf(stderr, "Can't find output size\n");
    return 0;
}

@@ -650,11 +582,16 @@

float *network_predict(network net, float *input)
{
    #ifdef GPU
#ifdef GPU
    if(gpu_index >= 0)  return network_predict_gpu(net, input);
    #endif
#endif

    forward_network(net, input, 0, 0);
    network_state state;
    state.input = input;
    state.truth = 0;
    state.train = 0;
    state.delta = 0;
    forward_network(net, state);
    float *out = get_network_output(net);
    return out;
}