>
> Hello guys, I've been working with content-based image retrieval and I want
> to use some visual descriptors in my thesis. Does any of you know some
> open-source implementation of these algorithms? I mean, something I could
> use in Python (with scikits.learn) or even C code which I could create a
> binding effortlessly.
You could try the SIFT descriptors from the VLFeat toolbox:
http://www.vlfeat.org/. These are C functions but should be easy to bind, if
not there are some python bindings available:
https://github.com/mmmikael/vlfeat. I've used the matlab interface to these
descriptors and they work very well.
There is a C implementation of HOG features available as part of this
toolbox:
http://www.eecs.berkeley.edu/~lbourdev/poselets/. This is a mex function
(the matlab C interface) however it's simple to adapt and run in python. You
just have to make sure that your numpy array is order 'F'. Easily achieved
by:
myarray = myarray.copy(dtype=float32, order = 'F')
Hope this helps,
Martin
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2d-oct
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
