How to have a faster sliding window algorithm ?

asked 2013-04-18 02:53:44 -0600

sub_o
175 ●1 ●14

I'm running a sliding window algorithm that extract features from the patch and feed it into my trained SVM. The sliding window itself is naive brute force sliding (i.e. sliding from 0 to max y, and 0 to max x)

The feature extraction algorithm is pretty fast, it took around 900 ms to slide and extract features from the entire image.

However passing the features into svm.predict( features, true ) slows it down to around 40 secs.

Is there any algorithm that can speed up the whole process ?

edit retag flag offensive close merge delete

add a comment

1 answer

Sort by » oldest newest most voted

answered 2013-04-19 09:55:33 -0600

Guanta
6736 ●6 ●25 ●79

Cascade classifiers are designed to reject false positives efficiently, maybe you want to try them out. Otherwise I have only the idea to use a multi-scale approach, maybe learn an SVM for larger windows and if there was a possible hit, pass it to another SVM, but I don't think it will be much faster...