Opencv vj级联检测器如何确定检测得分detection score在FDDB数据集测试

Opencv vj检测器如何确定detection score
从原理上来说vj检测器无法给出很符合的detection score，但是可以根据图像在a.哪个stage被拒；b.在被拒stage的重合窗口数；c.被拒stage返回的该stage level_weight，来得到detection score。FDDB数据集官网对这个问题就行了解答，采用a和c计算的方式。
但是opencv的detectMultiScale函数只返回最终阶段的reject_level和 level_weight，也就是说返回的reject_level就是vj检测器的层数，前面层被拒的图像没有保存
detectMultiScale函数说明：
This function allows you to retrieve the final stage decision certainty of classification.
For this, one needs to set outputRejectLevels on true and provide the rejectLevels and levelWeights parameter.
For each resulting detection, levelWeights will then contain the certainty of classification at the final stage.
This value can then be used to separate strong from weaker classifications.
FDDB数据集说明：

Q. How do you compute the detection score for OpenCV Viola-Jones detector?

In the OpenCV implementation, stage_sum is computed and compared against the i stage_threshold for each stage to accept/reject a candidate window. We define the detection score for a candidate window as
Kstage_when_rejected + stage_sum_for_stage_when_rejected.
If a window is accepted by the cascade, we just
Klast_stage + stage_sum_for_last_stage.
Choosing K as a large value e.g., 1000, we ensure that windows rejected at stage i have higher score than those rejected at stage i-1.
Q. I use 0/1 confidence scores for the face detections. The evaluation code only gave me two points on the ROC curve. Did I do something wrong?

The ROC curve is computed by varying the threshold over the confidence score. When you have {0,1} confidence score, there are only two points on the ROC curve. If you want a more detailed ROC curve, you would have to provide this confidence score as a real valued score.