YOLOv3计算recall和mAP

yolov3计算mAP有两种方法，第一种是使用faster rcnn中的voc_eval.py进行计算，另一种是通过修改yolov3中的代码进行计算。相比较而言第一种方法简单一些。

方法一：使用voc_eval.py进行计算

在yolov3中运行vaild命令进行测试。vaild命令执行的代码如下

void validate_detector(char *datacfg, char *cfgfile, char *weightfile, char *outfile)   
{
    int j;
    list *options = read_data_cfg(datacfg);
    char *valid_images = option_find_str(options, "valid", "data/train.list");
    char *name_list = option_find_str(options, "names", "data/names.list");
    char *prefix = option_find_str(options, "results", "results");
    char **names = get_labels(name_list);
  
//中间省略了部分代码

    } else {
        if(!outfile) outfile = "comp4_det_test_";
        fps = calloc(classes, sizeof(FILE *));
        for(j = 0; j < classes; ++j){
            snprintf(buff, 1024, "%s/%s%s.txt", prefix, outfile, names[j]);
            fps[j] = fopen(buff, "w");
        }
    }

从代码中可以看出，程序首先会根据你的数据集的类别创建对应的txt文件。创建的txt文件用来保存每一个类别检测到的目标框。txt文件创建的位置为 prefix + outfile + names[j], 其中prefix由data文件中的results指定， outfile默认为“comp4_det_test_”。

运行完成后，在对应的位置会看到生成的文件

我这里有6类，所以就生成了6个文件。有了这些文件后，就可以了使用voc_eval.py进行计算了。具体计算方法参见我的另一篇博客https://blog.csdn.net/sihaiyinan/article/details/89417963。

 

方法二：修改源代码进行计算

Windows版本下的darknet有计算mAP和recall的代码，把这部分代码稍加修改放到ubuntu版本的darknet中就可以了直接进行recall和mAP的计算了。代码修改后别忘了重新make一下。

1. 计算recall

在detector.c文件中有计算recall的函数，不过不能直接使用。直接把下面的代码替换成原来的validate_detector_recall函数即可。recall计算命令为

./darknet detector recall PATH/TO/*.data PATH/TO/*.cfg PATH/TO/*.weights

void validate_detector_recall(char *datacfg, char *cfgfile, char *weightfile)
{
    network *net = load_network(cfgfile, weightfile, 0);  //加载网络
    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));

    list *options = read_data_cfg(datacfg);   //读取data文件
    char *valid_images = option_find_str(options, "valid", "data/train.list");  //读取验证图像
    list *plist = get_paths(valid_images);
    char **paths = (char **)list_to_array(plist);

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

    int j, k;

    int m = plist->size;  //验证图像的数量
    int i=0;

    float thresh = .001;
    float iou_thresh = .5;
    float nms = .4;   

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

    
    for (i = 0; i < m; ++i) {
        char *path = paths[i];          //the i-th image's path
        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, sized.w, sized.h, thresh, .5, 0, 1, &nboxes);
        if (nms) do_nms_obj(dets, nboxes, 1, nms);  //抑制局部非最大

        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);

        int num_labels = 0;
        box_label *truth = read_boxes(labelpath, &num_labels);   //read the normalized image data in 'labels' folder
        
        for (k = 0; k < nboxes; ++k) {
            if (dets[k].objectness > thresh) {
                ++proposals;
            }
        }


        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 < nboxes; ++k) {                             //many boxs detected about the one object in the image, select the best box
                float iou = box_iou(dets[k].bbox, t);                  //compute IoU
                if (dets[k].objectness > thresh && iou > best_iou) {
                    best_iou = iou;
                }
            }
            avg_iou += best_iou;                 
            if (best_iou > iou_thresh) {        
                ++correct;
            }
        }
        //fprintf(stderr, " - %s\n - %s\n ", paths[i], lab