YOLO源码（Darknet源码）解读（yolo.c）

yolo.c

#include "darknet.h"

// 20个类别定义
char *voc_names[] = {"aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"};

void train_yolo(char *cfgfile, char *weightfile)
{
    char *train_images = "/data/voc/train.txt";
    char *backup_directory = "/home/pjreddie/backup/";

    // 初始化随机数生成器
    srand(time(0));

    // 从完整路径中截取出 文件名(不包含.后缀名)
    char *base = basecfg(cfgfile);
    printf("%s\n", base);

    // 根据cfg文件和weight文件加载网络
    network *net = load_network(cfgfile, weightfile, 0);
    printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", net->learning_rate, net->momentum, net->decay);

    int imgs = net->batch * net->subdivisions;
    int i = *net->seen / imgs;
    data train, buffer;
    float avg_loss = -1;

    layer l = net->layers[net->n - 1];

    int side = l.side;
    int classes = l.classes;
    float jitter = l.jitter;

    // 读取图像列表，并存储到二维字符串数组
    list *plist = get_paths(train_images);
    //int N = plist->size;
    char **paths = (char **)list_to_array(plist);

    load_args args = {0};
    args.w = net->w;
    args.h = net->h;
    args.paths = paths;
    args.n = imgs;
    args.m = plist->size;
    args.classes = classes;
    args.jitter = jitter;
    args.num_boxes = side;
    args.d = &buffer;
    args.type = REGION_DATA;

    args.angle = net->angle;
    args.exposure = net->exposure;
    args.saturation = net->saturation;
    args.hue = net->hue;

    // 加载图像，开始训练
    pthread_t load_thread = load_data_in_thread(args);
    clock_t time;
    //while(i*imgs < N*120){
    while(get_current_batch(net) < net->max_batches){
        i += 1;
        time=clock();
        pthread_join(load_thread, 0);
        train = buffer;
        load_thread = load_data_in_thread(args);

        printf("Loaded: %lf seconds\n", sec(clock()-time));

        time=clock();
        float loss = train_network(net, train);
        if (avg_loss < 0) avg_loss = loss;
        avg_loss = avg_loss*.9 + loss*.1;

        printf("%d: %f, %f avg, %f rate, %lf seconds, %d images\n", i, loss, avg_loss, get_current_rate(net), sec(clock()-time), i*imgs);

        // 存储迭代过程中的weights
        if(i%1000==0 || (i < 1000 && i%100 == 0)){
            char buff[256];
            sprintf(buff, "%s/%s_%d.weights", backup_directory, base, i);
            save_weights(net, buff);
        }

        free_data(train);
    }

    // 训练结束，存储weights
    char buff[256];
    sprintf(buff, "%s/%s_final.weights", backup_directory, base);
    save_weights(net, buff);
}

// 将检测的boxes结果写入文件
void print_yolo_detections(FILE **fps, char *id, int total, int classes, int w, int h, detection *dets)
{
    int i, j;
    for(i = 0; i < total; ++i){
        float xmin = dets[i].bbox.x - dets[i].bbox.w/2.;
        float xmax = dets[i].bbox.x + dets[i].bbox.w/2.;
        float ymin = dets[i].bbox.y - dets[i].bbox.h/2.;
        float ymax = dets[i].bbox.y + dets[i].bbox.h/2.;

        if (xmin < 0) xmin = 0;
        if (ymin < 0) ymin = 0;
        if (xmax > w) xmax = w;
        if (ymax > h) ymax = h;

        for(j = 0; j < classes; ++j){
            if (dets[i].prob[j]) fprintf(fps[j], "%s %f %f %f %f %f\n", id, dets[i].prob[j],
                    xmin, ymin, xmax, ymax);
        }
    }
}

// 在验证集上检测物体
void validate_yolo(char *cfg, char *weights)
{
    network *net = load_network(cfg, weights, 0);
    set_batch_network(net, 1);

    fprintf(stderr, "Learning Rate: %g, Momentum: %g, Decay: %g\n", net->learning_rate, net->momentum, net->decay);
    srand(time(0));

    char *base = "results/comp4_det_test_";
    //list *plist = get_paths("data/voc.2007.test");
    list *plist = get_paths("/home/pjreddie/data/voc/2007_test.txt");
    //list *plist = get_paths("data/voc.2012.test");
    char **paths = (char **)list_to_array(plist);

    layer l = net->layers[net->n-1];
    int classes = l.classes;

    // 一个label创建一个文件，存储检测结果
    int j;
    FILE **fps = calloc(classes, sizeof(FILE *));
    for(j = 0; j < classes; ++j){
        char buff[1024];
        snprintf(buff, 1024, "%s%s.txt", base, voc_names[j]);
        fps[j] = fopen(buff, "w");
    }

    int m = plist->size;
    int i=0;
    int t;

    float thresh = .001;
    int nms = 1;
    float iou_thresh = .5;

    int nthreads = 8;
    image *val = calloc(nthreads, sizeof(image));
    image *val_resized = calloc(nthreads, sizeof(image));
    image *buf = calloc(nthreads, sizeof(image));
    image *buf_resized = calloc(nthreads, sizeof(image));
    pthread_t *thr = calloc(nthreads, sizeof(pthread_t));

    load_args args = {0};
    args.w = net->w;
    args.h = net->h;
    args.type = IMAGE_DATA;

    // 创建线程，读取图像，对应图像id [0, nthreads - 1]
    for(t = 0; t < nthreads; ++t){
        args.path = paths[i+t];
        args.im = &buf[t];
        args.resized = &buf_resized[t];
        thr[t] = load_data_in_thread(args);
    }

    time_t start = time(0);
    for(i = nthreads; i < m+nthreads; i += nthreads){
        fprintf(stderr, "%d\n", i);

        // 等待线程结束，对应图像id [i - nthreads, i - 1]
        for(t = 0; t < nthreads && i+t-nthreads < m; ++t){
            pthread_join(thr[t], 0);
            val[t] = buf[t];
            val_resized[t] = buf_resized[t];
        }

        // 创建线程，读取图像，对应图像id [i, min(m - 1, i + nthread - 1)]
        for(t = 0; t < nthreads && i+t < m; ++t){
            args.path = paths[i+t];
            args.im = &buf[t];
            args.resized = &buf_resized[t];
            thr[t] = load_data_in_thread(args);
        }

        // 检测图像中的物体，对应图像id [i - nthreads, i - 1]
        for(t = 0; t < nthreads && i+t-nthreads < m; ++t){
            // 获取图像文件名
            char *path = paths[i+t-nthreads];
            char *id = basecfg(path);

            // 检测
            float *X = val_resized[t].data;
            network_predict(net, X);

            // 获取bboxes
            int w = val[t].w;
            int h = val[t].h;
            int nboxes = 0;
            detection *dets = get_network_boxes(net, w, h, thresh, 0, 0, 0, &nboxes);

            // nms
            if (nms) do_nms_sort(dets, l.side*l.side*l.n, classes, iou_thresh);

            // print
            print_yolo_detections(fps, id, l.side*l.side*l.n, classes, w, h, dets);

            free_detections(dets, nboxes);
            free(id);
            free_image(val[t]);
            free_image(val_resized[t]);
        }
    }

    fprintf(stderr, "Total Detection Time: %f Seconds\n", (double)(time(0) - start));
}

// 在验证集上计算 bbox objectness 的 recall
void validate_yolo_recall(char *cfg, char *weights)
{
    network *net = load_network(cfg, weights, 0);
    set_batch_network(net, 1);
    fprintf(stderr, "Learning Rate: %g, Momentum: %g, Decay: %g\n", net->learning_rate, net->momentum, net->decay);
    srand(time(0));

    char *base = "results/comp4_det_test_";
    list *plist = get_paths("data/voc.2007.test");
    char **paths = (char **)list_to_array(plist);

    layer l = net->layers[net->n-1];
    int classes = l.classes;
    int side = l.side;

    int j, k;
    FILE **fps = calloc(classes, sizeof(FILE *));
    for(j = 0; j < classes; ++j){
        char buff[1024];
        snprintf(buff, 1024, "%s%s.txt", base, voc_names[j]);
        fps[j] = fopen(buff, "w");
    }

    int m = plist->size;
    int i=0;

    float thresh = .001;
    float iou_thresh = .5;
    float nms = 0;

    int total = 0;
    int correct = 0;
    int proposals = 0;
    float avg_iou = 0;

    for(i = 0; i < m; ++i){
        char *path = paths[i];
        image orig = load_image_color(path, 0, 0);
        image sized = resize_image(orig, net->w, net->h);
        char *id = basecfg(path);
        network_predict(net, sized.data);

        int nboxes = 0;
        detection *dets = get_network_boxes(net, orig.w, orig.h, thresh, 0, 0, 1, &nboxes);
        if (nms) do_nms_obj(dets, side*side*l.n, 1, nms);

        // 根据 image path 和替换规则，构建 label path
        char labelpath[4096];
        find_replace(path, "images", "labels", labelpath);
        find_replace(labelpath, "JPEGImages", "labels", labelpath);
        find_replace(labelpath, ".jpg", ".txt", labelpath);
        find_replace(labelpath, ".JPEG", ".txt", labelpath);

        // 读取 ground truth
        int num_labels = 0;
        box_label *truth = read_boxes(labelpath, &num_labels);

        // 统计 proposal 个数
        for(k = 0; k < side*side*l.n; ++k){
            if(dets[k].objectness > thresh){
                ++proposals;
            }
        }

        // 给每一个 ground truth 匹配 iou 最大的 bbox
        for (j = 0; j < num_labels; ++j) {
            ++total;

            box t = {truth[j].x, truth[j].y, truth[j].w, truth[j].h};

            float best_iou = 0;
            for(k = 0; k < side*side*l.n; ++k){
                float iou = box_iou(dets[k].bbox, t);
                if(dets[k].objectness > thresh && iou > best_iou){
                    best_iou = iou;
                }
            }

            if(best_iou > iou_thresh){
                ++correct;
            }

            avg_iou += best_iou;
        }

        fprintf(stderr, "%5d %5d %5d\tRPs/Img: %.2f\tIOU: %.2f%%\tRecall:%.2f%%\n", i, correct, total, (float)proposals/(i+1), avg_iou*100/total, 100.*correct/total);

        free_detections(dets, nboxes);
        free(id);
        free_image(orig);
        free_image(sized);
    }
}

// 检测一张图像或多多张图像
void test_yolo(char *cfgfile, char *weightfile, char *filename, float thresh)
{
    image **alphabet = load_alphabet();
    network *net = load_network(cfgfile, weightfile, 0);
    layer l = net->layers[net->n-1];
    set_batch_network(net, 1);

    srand(2222222);
    clock_t time;
    char buff[256];
    char *input = buff;
    float nms=.4;

    while(1){
        // 获取图像路径
        if(filename){
            strncpy(input, filename, 256);；
        } else {
            printf("Enter Image Path: ");
            fflush(stdout);
            input = fgets(input, 256, stdin);
            if(!input) return;
            strtok(input, "\n");
        }

        image im = load_image_color(input,0,0);
        image sized = resize_image(im, net->w, net->h);
        float *X = sized.data;
        time=clock();
        network_predict(net, X);
        printf("%s: Predicted in %f seconds.\n", input, sec(clock()-time));

        int nboxes = 0;
        detection *dets = get_network_boxes(net, 1, 1, thresh, 0, 0, 0, &nboxes);
        if (nms) do_nms_sort(dets, l.side*l.side*l.n, l.classes, nms);

        // 在图像上画出 bboxes，显示出来，并保存
        draw_detections(im, dets, l.side*l.side*l.n, thresh, voc_names, alphabet, 20);
        save_image(im, "predictions");
        show_image(im, "predictions");

        free_detections(dets, nboxes);
        free_image(im);
        free_image(sized);
#ifdef OPENCV
        cvWaitKey(0);
        cvDestroyAllWindows();
#endif
        if (filename) break;
    }
}

// YOLO入口，根据入参区分 train/valid/test
void run_yolo(int argc, char **argv)
{
    char *prefix = find_char_arg(argc, argv, "-prefix", 0);
    float thresh = find_float_arg(argc, argv, "-thresh", .2);
    int cam_index = find_int_arg(argc, argv, "-c", 0);
    int frame_skip = find_int_arg(argc, argv, "-s", 0);
    if(argc < 4){
        fprintf(stderr, "usage: %s %s [train/test/valid] [cfg] [weights (optional)]\n", argv[0], argv[1]);
        return;
    }

    int avg = find_int_arg(argc, argv, "-avg", 1);
    char *cfg = argv[3];
    char *weights = (argc > 4) ? argv[4] : 0;
    char *filename = (argc > 5) ? argv[5]: 0;
    if(0==strcmp(argv[2], "test")) test_yolo(cfg, weights, filename, thresh);
    else if(0==strcmp(argv[2], "train")) train_yolo(cfg, weights);
    else if(0==strcmp(argv[2], "valid")) validate_yolo(cfg, weights);
    else if(0==strcmp(argv[2], "recall")) validate_yolo_recall(cfg, weights);
    else if(0==strcmp(argv[2], "demo")) demo(cfg, weights, thresh, cam_index, filename, voc_names, 20, frame_skip, prefix, avg, .5, 0,0,0,0);
}