cleaned up data parsing a lot. probably nothing broken?

src/cnn.c

@@ -269,13 +269,6 @@ void test_parser()
     save_network(net, "cfg/trained_imagenet_smaller.cfg");
 }
 
-void test_data()
-{
-	char *labels[] = {"cat","dog"};
-	data train = load_data_image_pathfile_random("train_paths.txt", 101,labels, 2, 300, 400);
-	free_data(train);
-}
-
 void train_asirra()
 {
     network net = parse_network_cfg("cfg/imagenet.cfg");
@@ -284,11 +277,18 @@ void train_asirra()
     srand(2222222);
     int i = 0;
     char *labels[] = {"cat","dog"};
+
+    list *plist = get_paths("data/assira/train.list");
+    char **paths = (char **)list_to_array(plist);
+    int m = plist->size;
+    free_list(plist);
+
     clock_t time;
+
     while(1){
         i += 1;
         time=clock();
-		data train = load_data_image_pathfile_random("data/assira/train.list", imgs*net.batch, labels, 2, 256, 256);
+        data train = load_data_random(imgs*net.batch, paths, m, labels, 2, 256, 256);
         normalize_data_rows(train);
         printf("Loaded: %lf seconds\n", sec(clock()-time));
         time=clock();
@@ -305,11 +305,48 @@ void train_asirra()
     }
 }
 
+void train_detection_net()
+{
+    float avg_loss = 1;
+    //network net = parse_network_cfg("/home/pjreddie/imagenet_backup/alexnet_1270.cfg");
+    network net = parse_network_cfg("cfg/detnet.cfg");
+    printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", net.learning_rate, net.momentum, net.decay);
+    int imgs = 1000/net.batch+1;
+    srand(time(0));
+    int i = 0;
+    char **labels = get_labels("/home/pjreddie/data/imagenet/cls.labels.list");
+    list *plist = get_paths("/data/imagenet/cls.train.list");
+    char **paths = (char **)list_to_array(plist);
+    printf("%d\n", plist->size);
+    clock_t time;
+    while(1){
+        i += 1;
+        time=clock();
+        data train = load_data_random(imgs*net.batch, paths, plist->size, labels, 1000, 256, 256);
+        //translate_data_rows(train, -144);
+        normalize_data_rows(train);
+        printf("Loaded: %lf seconds\n", sec(clock()-time));
+        time=clock();
+#ifdef GPU
+        float loss = train_network_data_gpu(net, train, imgs);
+        avg_loss = avg_loss*.9 + loss*.1;
+        printf("%d: %f, %f avg, %lf seconds, %d images\n", i, loss, avg_loss, sec(clock()-time), i*imgs*net.batch);
+#endif
+        free_data(train);
+        if(i%10==0){
+            char buff[256];
+            sprintf(buff, "/home/pjreddie/imagenet_backup/imagenet_%d.cfg", i);
+            save_network(net, buff);
+        }
+    }
+}
+
+
 void train_imagenet()
 {
     float avg_loss = 1;
     //network net = parse_network_cfg("/home/pjreddie/imagenet_backup/alexnet_1270.cfg");
-	network net = parse_network_cfg("cfg/imagenet.cfg");
+    network net = parse_network_cfg("cfg/alexnet.cfg");
     printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", net.learning_rate, net.momentum, net.decay);
     int imgs = 1000/net.batch+1;
     srand(time(0));
@@ -348,14 +385,21 @@ void validate_imagenet(char *filename)
     srand(time(0));
 
     char **labels = get_labels("/home/pjreddie/data/imagenet/cls.val.labels.list");
-    char *path = "/home/pjreddie/data/imagenet/cls.val.list";
+
+    list *plist = get_paths("/home/pjreddie/data/imagenet/cls.val.list");
+    char **paths = (char **)list_to_array(plist);
+    int m = plist->size;
+    free_list(plist);
 
     clock_t time;
     float avg_acc = 0;
     int splits = 50;
+
     for(i = 0; i < splits; ++i){
         time=clock();
-        data val = load_data_image_pathfile_part(path, i, splits, labels, 1000, 256, 256);
+        char **part = paths+(i*m/splits);
+        int num = (i+1)*m/splits - i*m/splits;
+        data val = load_data(part, num, labels, 1000, 256, 256);
         normalize_data_rows(val);
         printf("Loaded: %d images in %lf seconds\n", val.X.rows, sec(clock()-time));
         time=clock();
@@ -368,35 +412,6 @@ void validate_imagenet(char *filename)
     }
 }
 
-void train_imagenet_small()
-{
-	network net = parse_network_cfg("cfg/imagenet_small.cfg");
-    printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", net.learning_rate, net.momentum, net.decay);
-    int imgs=1;
-    srand(111222);
-    int i = 0;
-    char **labels = get_labels("/home/pjreddie/data/imagenet/cls.labels.list");
-    list *plist = get_paths("/data/imagenet/cls.train.list");
-    char **paths = (char **)list_to_array(plist);
-    printf("%d\n", plist->size);
-    clock_t time;
-
-    i += 1;
-    time=clock();
-    data train = load_data_random(imgs*net.batch, paths, plist->size, labels, 1000, 256, 256);
-    normalize_data_rows(train);
-    printf("Loaded: %lf seconds\n", sec(clock()-time));
-    time=clock();
-#ifdef GPU
-    float loss = train_network_data_gpu(net, train, imgs);
-    printf("%d: %f, %lf seconds, %d images\n", i, loss, sec(clock()-time), i*imgs*net.batch);
-#endif
-    free_data(train);
-    char buff[256];
-    sprintf(buff, "/home/pjreddie/imagenet_backup/imagenet_backup_slower_larger_%d.cfg", i);
-    save_network(net, buff);
-}
-
 void test_imagenet()
 {
     network net = parse_network_cfg("cfg/imagenet_test.cfg");
@@ -431,32 +446,6 @@ void test_visualize(char *filename)
     visualize_network(net);
     cvWaitKey(0);
 }
-void test_full()
-{
-    network net = parse_network_cfg("cfg/backup_1300.cfg");
-    srand(2222222);
-    int i,j;
-    int total = 100;
-    char *labels[] = {"cat","dog"};
-    FILE *fp = fopen("preds.txt","w");
-    for(i = 0; i < total; ++i){
-        visualize_network(net);
-        cvWaitKey(100);
-        data test = load_data_image_pathfile_part("data/assira/test.list", i, total, labels, 2, 256, 256);
-        image im = float_to_image(256, 256, 3,test.X.vals[0]);
-        show_image(im, "input");
-        cvWaitKey(100);
-        normalize_data_rows(test);
-        for(j = 0; j < test.X.rows; ++j){
-            float *x = test.X.vals[j];
-            forward_network(net, x, 0, 0);
-            int class = get_predicted_class_network(net);
-            fprintf(fp, "%d\n", class);
-        }
-        free_data(test);
-    }
-    fclose(fp);
-}
 
 void test_cifar10()
 {
@@ -675,88 +664,74 @@ void flip_network()
     save_network(net, "cfg/voc_imagenet_rev.cfg");
 }
 
-void tune_VOC()
+
+void visualize_cat()
 {
-    network net = parse_network_cfg("cfg/voc_start.cfg");
+    network net = parse_network_cfg("cfg/voc_imagenet.cfg");
-    srand(2222222);
+    image im = load_image("data/cat.png", 0, 0);
-    int i = 20;
+    printf("Processing %dx%d image\n", im.h, im.w);
-    char *labels[] = {"aeroplane","bicycle","bird","boat","bottle","bus","car","cat","chair","cow","diningtable","dog","horse","motorbike","person","pottedplant","sheep","sofa","train","tvmonitor"};
-    float lr = .000005;
-    float momentum = .9;
-    float decay = 0.0001;
-    while(i++ < 1000 || 1){
-        data train = load_data_image_pathfile_random("/home/pjreddie/VOC2012/trainval_paths.txt", 10, labels, 20, 256, 256);
-
-        image im = float_to_image(256, 256, 3,train.X.vals[0]);
-        show_image(im, "input");
-        visualize_network(net);
-        cvWaitKey(100);
-
-        translate_data_rows(train, -144);
-        clock_t start = clock(), end;
-        float loss = train_network_sgd(net, train, 10);
-        end = clock();
-        printf("%d: %f, Time: %lf seconds, LR: %f, Momentum: %f, Decay: %f\n", i, loss, (float)(end-start)/CLOCKS_PER_SEC, lr, momentum, decay);
-        free_data(train);
-        /*
-           if(i%10==0){
-           char buff[256];
-           sprintf(buff, "/home/pjreddie/voc_cfg/voc_ramp_%d.cfg", i);
-           save_network(net, buff);
-           }
-         */
-        //lr *= .99;
-    }
-}
-
-int voc_size(int x)
-{
-    x = x-1+3;
-    x = x-1+3;
-    x = x-1+3;
-    x = (x-1)*2+1;
-    x = x-1+5;
-    x = (x-1)*2+1;
-    x = (x-1)*4+11;
-    return x;
-}
-
-image features_output_size(network net, IplImage *src, int outh, int outw)
-{
-    int h = voc_size(outh);
-    int w = voc_size(outw);
-    fprintf(stderr, "%d %d\n", h, w);
-
-    IplImage *sized = cvCreateImage(cvSize(w,h), src->depth, src->nChannels);
-    cvResize(src, sized, CV_INTER_LINEAR);
-    image im = ipl_to_image(sized);
-    //normalize_array(im.data, im.h*im.w*im.c);
-    translate_image(im, -144);
     resize_network(net, im.h, im.w, im.c);
     forward_network(net, im.data, 0, 0);
-    image out = get_network_image(net);
+
-    free_image(im);
+    visualize_network(net);
-    cvReleaseImage(&sized);
+    cvWaitKey(0);
-    return copy_image(out);
 }
 
-void features_VOC_image_size(char *image_path, int h, int w)
+
+void test_gpu_net()
 {
-    int j;
+    srand(222222);
-    network net = parse_network_cfg("cfg/voc_imagenet.cfg");
+    network net = parse_network_cfg("cfg/nist.cfg");
-    fprintf(stderr, "%s\n", image_path);
+    data train = load_categorical_data_csv("data/mnist/mnist_train.csv", 0, 10);
+    data test = load_categorical_data_csv("data/mnist/mnist_test.csv",0,10);
+    translate_data_rows(train, -144);
+    translate_data_rows(test, -144);
+    int count = 0;
+    int iters = 1000/net.batch;
+    while(++count <= 5){
+        clock_t start = clock(), end;
+        float loss = train_network_sgd(net, train, iters);
+        end = clock();
+        float test_acc = network_accuracy(net, test);
+        printf("%d: Loss: %f, Test Acc: %f, Time: %lf seconds, LR: %f, Momentum: %f, Decay: %f\n", count, loss, test_acc,(float)(end-start)/CLOCKS_PER_SEC, net.learning_rate, net.momentum, net.decay);
+    }
+#ifdef GPU
+    count = 0;
+    srand(222222);
+    net = parse_network_cfg("cfg/nist.cfg");
+    while(++count <= 5){
+        clock_t start = clock(), end;
+        float loss = train_network_sgd_gpu(net, train, iters);
+        end = clock();
+        float test_acc = network_accuracy(net, test);
+        printf("%d: Loss: %f, Test Acc: %f, Time: %lf seconds, LR: %f, Momentum: %f, Decay: %f\n", count, loss, test_acc,(float)(end-start)/CLOCKS_PER_SEC, net.learning_rate, net.momentum, net.decay);
+    }
+#endif
+}
 
-    IplImage* src = 0;
+
-    if( (src = cvLoadImage(image_path,-1)) == 0 ) file_error(image_path);
+int main(int argc, char *argv[])
-    image out = features_output_size(net, src, h, w);
+{
-    for(j = 0; j < out.c*out.h*out.w; ++j){
+    if(argc < 2){
-        if(j != 0) printf(",");
+        fprintf(stderr, "usage: %s <function>\n", argv[0]);
-        printf("%g", out.data[j]);
+        return 0;
     }
-    printf("\n");
+    if(0==strcmp(argv[1], "train")) train_imagenet();
-    free_image(out);
+    else if(0==strcmp(argv[1], "asirra")) train_asirra();
-    cvReleaseImage(&src);
+    else if(0==strcmp(argv[1], "nist")) train_nist();
+    else if(0==strcmp(argv[1], "test_correct")) test_gpu_net();
+    else if(0==strcmp(argv[1], "test")) test_imagenet();
+    else if(0==strcmp(argv[1], "visualize")) test_visualize(argv[2]);
+    else if(0==strcmp(argv[1], "valid")) validate_imagenet(argv[2]);
+#ifdef GPU
+    else if(0==strcmp(argv[1], "test_gpu")) test_gpu_blas();
+#endif
+    test_parser();
+    fprintf(stderr, "Success!\n");
+    return 0;
 }
+
+/*
 void visualize_imagenet_topk(char *filename)
 {
     int i,j,k,l;
@@ -873,19 +848,6 @@ void visualize_imagenet_features(char *filename)
     }
     cvWaitKey(0);
 }
-
-void visualize_cat()
-{
-    network net = parse_network_cfg("cfg/voc_imagenet.cfg");
-    image im = load_image("data/cat.png", 0, 0);
-    printf("Processing %dx%d image\n", im.h, im.w);
-    resize_network(net, im.h, im.w, im.c);
-    forward_network(net, im.data, 0, 0);
-
-    visualize_network(net);
-    cvWaitKey(0);
-}
-
 void features_VOC_image(char *image_file, char *image_dir, char *out_dir, int flip, int interval)
 {
     int i,j;
@@ -992,57 +954,4 @@ void test_distribution()
     cvWaitKey(0);
     cvWaitKey(0);
 }
-
+*/
-void test_gpu_net()
-{
-    srand(222222);
-    network net = parse_network_cfg("cfg/nist.cfg");
-    data train = load_categorical_data_csv("data/mnist/mnist_train.csv", 0, 10);
-    data test = load_categorical_data_csv("data/mnist/mnist_test.csv",0,10);
-    translate_data_rows(train, -144);
-    translate_data_rows(test, -144);
-    int count = 0;
-    int iters = 1000/net.batch;
-    while(++count <= 5){
-        clock_t start = clock(), end;
-        float loss = train_network_sgd(net, train, iters);
-        end = clock();
-        float test_acc = network_accuracy(net, test);
-        printf("%d: Loss: %f, Test Acc: %f, Time: %lf seconds, LR: %f, Momentum: %f, Decay: %f\n", count, loss, test_acc,(float)(end-start)/CLOCKS_PER_SEC, net.learning_rate, net.momentum, net.decay);
-    }
-    #ifdef GPU
-    count = 0;
-    srand(222222);
-    net = parse_network_cfg("cfg/nist.cfg");
-    while(++count <= 5){
-        clock_t start = clock(), end;
-        float loss = train_network_sgd_gpu(net, train, iters);
-        end = clock();
-        float test_acc = network_accuracy(net, test);
-        printf("%d: Loss: %f, Test Acc: %f, Time: %lf seconds, LR: %f, Momentum: %f, Decay: %f\n", count, loss, test_acc,(float)(end-start)/CLOCKS_PER_SEC, net.learning_rate, net.momentum, net.decay);
-    }
-    #endif
-}
-
-
-int main(int argc, char *argv[])
-{
-    if(argc < 2){
-        fprintf(stderr, "usage: %s <function>\n", argv[0]);
-        return 0;
-    }
-    if(0==strcmp(argv[1], "train")) train_imagenet();
-    else if(0==strcmp(argv[1], "asirra")) train_asirra();
-    else if(0==strcmp(argv[1], "nist")) train_nist();
-    else if(0==strcmp(argv[1], "train_small")) train_imagenet_small();
-    else if(0==strcmp(argv[1], "test_correct")) test_gpu_net();
-    else if(0==strcmp(argv[1], "test")) test_imagenet();
-    else if(0==strcmp(argv[1], "visualize")) test_visualize(argv[2]);
-    else if(0==strcmp(argv[1], "valid")) validate_imagenet(argv[2]);
-    #ifdef GPU
-    else if(0==strcmp(argv[1], "test_gpu")) test_gpu_blas();
-    #endif
-    test_parser();
-    fprintf(stderr, "Success!\n");
-    return 0;
-}

src/data.c

@@ -19,10 +19,28 @@ list *get_paths(char *filename)
     return lines;
 }
 
-void fill_truth_det(char *path, float *truth)
+void fill_truth_detection(char *path, float *truth, int height, int width, int num_height, int num_width, float scale)
 {
-    find_replace(path, "imgs", "det");
+    int box_height = height/num_height;
-    find_replace(path, ".JPEG", ".txt");
+    int box_width = width/num_width;
+    char *labelpath = find_replace(path, "imgs", "det");
+    labelpath = find_replace(labelpath, ".JPEG", ".txt");
+    FILE *file = fopen(labelpath, "r");
+    int x, y, h, w;
+    while(fscanf(file, "%d %d %d %d", &x, &y, &w, &h) == 4){
+        int i = x/box_width;
+        int j = y/box_height;
+        float dh = (float)(x%box_width)/box_height;
+        float dw = (float)(y%box_width)/box_width;
+        float sh = h/scale;
+        float sw = w/scale;
+        int index = (i+j*num_width)*5;
+        truth[index++] = 1;
+        truth[index++] = dh;
+        truth[index++] = dw;
+        truth[index++] = sh;
+        truth[index++] = sw;
+    }
 }
 
 void fill_truth(char *path, char **labels, int k, float *truth)
@@ -36,32 +54,52 @@ void fill_truth(char *path, char **labels, int k, float *truth)
     }
 }
 
-data load_data_image_paths(char **paths, int n, char **labels, int k, int h, int w)
+matrix load_image_paths(char **paths, int n, int h, int w)
 {
     int i;
-    data d;
+    matrix X;
-    d.shallow = 0;
+    X.rows = n;
-    d.X.rows = n;
+    X.vals = calloc(X.rows, sizeof(float*));
-    d.X.vals = calloc(d.X.rows, sizeof(float*));
+    X.cols = 0;
-    d.X.cols = 0;
-    d.y = make_matrix(n, k);
 
     for(i = 0; i < n; ++i){
         image im = load_image_color(paths[i], h, w);
-        d.X.vals[i] = im.data;
+        X.vals[i] = im.data;
-        d.X.cols = im.h*im.w*im.c;
+        X.cols = im.h*im.w*im.c;
     }
+    return X;
+}
+
+matrix load_labels_paths(char **paths, int n, char **labels, int k)
+{
+    matrix y = make_matrix(n, k);
+    int i;
     for(i = 0; i < n; ++i){
-        fill_truth(paths[i], labels, k, d.y.vals[i]);
+        fill_truth(paths[i], labels, k, y.vals[i]);
     }
-    return d;
+    return y;
+}
+
+matrix load_labels_detection(char **paths, int n, int height, int width, int num_height, int num_width, float scale)
+{
+    int k = num_height*num_width*5;
+    matrix y = make_matrix(n, k);
+    int i;
+    for(i = 0; i < n; ++i){
+        fill_truth_detection(paths[i], y.vals[i], height, width, num_height, num_width, scale);
+    }
+    return y;
 }
 
 data load_data_image_pathfile(char *filename, char **labels, int k, int h, int w)
 {
     list *plist = get_paths(filename);
     char **paths = (char **)list_to_array(plist);
-    data d = load_data_image_paths(paths, plist->size, labels, k, h, w);
+    int n = plist->size;
+    data d;
+    d.shallow = 0;
+    d.X = load_image_paths(paths, n, h, w);
+    d.y = load_labels_paths(paths, n, labels, k);
     free_list_contents(plist);
     free_list(plist);
     free(paths);
@@ -87,16 +125,29 @@ void free_data(data d)
     }
 }
 
-data load_data_image_pathfile_part(char *filename, int part, int total, char **labels, int k, int h, int w)
+data load_data_detection_random(int n, char **paths, int m, char **labels, int h, int w, int nh, int nw, float scale)
 {
-    list *plist = get_paths(filename);
+    char **random_paths = calloc(n, sizeof(char*));
-    char **paths = (char **)list_to_array(plist);
+    int i;
-    int start = part*plist->size/total;
+    for(i = 0; i < n; ++i){
-    int end = (part+1)*plist->size/total;
+        int index = rand()%m;
-    data d = load_data_image_paths(paths+start, end-start, labels, k, h, w);
+        random_paths[i] = paths[index];
-    free_list_contents(plist);
+        if(i == 0) printf("%s\n", paths[index]);
-    free_list(plist);
+    }
-    free(paths);
+    data d;
+    d.shallow = 0;
+    d.X = load_image_paths(random_paths, n, h, w);
+    d.y = load_labels_detection(random_paths, n, h, w, nh, nw, scale);
+    free(random_paths);
+    return d;
+}
+
+data load_data(char **paths, int n, char **labels, int k, int h, int w)
+{
+    data d;
+    d.shallow = 0;
+    d.X = load_image_paths(paths, n, h, w);
+    d.y = load_labels_paths(paths, n, labels, k);
     return d;
 }
 
@@ -109,26 +160,7 @@ data load_data_random(int n, char **paths, int m, char **labels, int k, int h, i
         random_paths[i] = paths[index];
         if(i == 0) printf("%s\n", paths[index]);
     }
-    data d = load_data_image_paths(random_paths, n, labels, k, h, w);
+    data d = load_data(random_paths, n, labels, k, h, w);
-    free(random_paths);
-    return d;
-}
-
-data load_data_image_pathfile_random(char *filename, int n, char **labels, int k, int h, int w)
-{
-    int i;
-    list *plist = get_paths(filename);
-    char **paths = (char **)list_to_array(plist);
-    char **random_paths = calloc(n, sizeof(char*));
-    for(i = 0; i < n; ++i){
-        int index = rand()%plist->size;
-        random_paths[i] = paths[index];
-        if(i == 0) printf("%s\n", paths[index]);
-    }
-    data d = load_data_image_paths(random_paths, n, labels, k, h, w);
-    free_list_contents(plist);
-    free_list(plist);
-    free(paths);
     free(random_paths);
     return d;
 }

src/data.h

@@ -12,12 +12,10 @@ typedef struct{
 
 
 void free_data(data d);
+data load_data(char **paths, int n, char **labels, int k, int h, int w);
 data load_data_random(int n, char **paths, int m, char **labels, int k, int h, int w);
+data load_data_detection_random(int n, char **paths, int m, char **labels, int h, int w, int nh, int nw, float scale);
 data load_data_image_pathfile(char *filename, char **labels, int k, int h, int w);
-data load_data_image_pathfile_part(char *filename, int part, int total, 
-                                    char **labels, int k, int h, int w);
-data load_data_image_pathfile_random(char *filename, int n, char **labels, 
-                                        int k, int h, int w);
 data load_cifar10_data(char *filename);
 data load_all_cifar10();
 list *get_paths(char *filename);