How to get TP,TN,FP,FN of TrainCasca​deObjectDe​tector?

2 次查看（过去 30 天）
matlab22
matlab22 2016-5-11
hello, i used TrainCascadeObjectDetector for training a classifier i want to plot a ROC curve to evalute detector, i need TP , TN , FP , FN, how to get them for each stage of the classification, for exemple this simple training showing for every stage just the pos and neg samples used:
Automatically setting ObjectTrainingSize to [ 32, 87 ]
Using at most 28 of 28 positive samples per stage
Using at most 56 negative samples per stage
Training stage 1 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 1: 1 seconds
Training stage 2 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 2: 0 seconds
Training stage 3 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 3: 1 seconds
Training stage 4 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 4: 1 seconds
Training stage 5 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 5: 1 seconds
Training stage 6 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 6: 1 seconds
Training stage 7 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 7: 1 seconds
Training stage 8 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 8: 5 seconds
Training stage 9 of 16
[........................................................................]
Used 28 positive and 56 negative samples
Time to train stage 9: 11 seconds
Training stage 10 of 16
[........................................................................]
Used 28 positive and 22 negative samples
Time to train stage 10: 9 seconds
Training stage 11 of 16
[........................................................................]
Used 28 positive and 22 negative samples
Time to train stage 11: 9 seconds
Training stage 12 of 16
[........................................................................]
Used 28 positive and 13 negative samples
Time to train stage 12: 9 seconds
Training stage 13 of 16
[........................................................................]
Used 28 positive and 7 negative samples
Time to train stage 13: 9 seconds
Training stage 14 of 16
[........................................................................]
Used 28 positive and 8 negative samples
Time to train stage 14: 10 seconds
.
.
.
In other words, how to keep track of these values TP TN FP FN for each stage?

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Statistics and Machine Learning Toolbox 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by