Why is opencv SVM learning much faster than libsvm 3.18?

I am comparing OpenCV SVM and libsvm 3.18 and with the same parameters and dataset, OpenCV svm training time is about 6-8x faster than libsvm 3.18. I even set maximum iterations in the libsvm code. He said that OpenCV SVM is based on libsvm, but I'm wondering what the difference is between them. Thanks to

+3

opencv machine-learning svm libsvm

Mickey shine 19 Aug '14 at 7:18

source to share

No one has answered this question yet

Check out similar questions:

4

OpenCV for Android - learning SVM with SURF descriptors

3

Big data training and testing in libsvm

3

training classification using libsvm

2

Feature selection SVM-recursive elimination function (SVM-RFE) with Libsvm, the precision result is worse than without feature selection, why?

1

Differences between decoding between Scikit-learn SVM and libsvm in Matlab

1

LibSVM for multilevel precision SVM

1

SVM LibSVM ignores function 1,3,5 when predicting

1

Parameter estimation for linear training on one SVM class via libsvm for n-grams

0

R e1071 trains faster than Libsvm

0

why the size of the model file trained by libsvm is much larger than the model file trained by opencv

All Articles