i am doing my project on face, age and gender recognition. i am using the LBP technique for the same. but, when i initially train the system for the first time, the XML file created is about 44MB, and this makes the further processes slow. could anyone tell me how big the XML file should actually be? also could you give me suggestions to improve the efficiency of the face recognition code?
The large file is also due to a missing dimensionality reduction in the implementation, which can be applied for Local Binary Patterns without a great loss of information. If you read up some literature on Local Binary Patterns, you'll notice there is something called Uniform Patterns. This is due to the LBP definition and its a-priori probabilities for building some patterns :
The Scholarpedia page on LBP states:
Another extension to the original operator is the definition of so called uniform patterns, which can be used to reduce the length of the feature vector and implement a simple rotation-invariant descriptor. This extension was inspired by the fact that some binary patterns occur more commonly in texture images than others. A local binary pattern is called uniform if the binary pattern contains at most two bitwise transitions from 0 to 1 or vice versa when the bit pattern is traversed circularly. For example, the patterns 00000000 (0 transitions), 01110000 (2 transitions) and 11001111 (2 transitions) are uniform whereas the patterns 11001001 (4 transitions) and 01010010 (6 transitions) are not. In the computation of the LBP labels, uniform patterns are used so that there is a separate label for each uniform pattern and all the non-uniform patterns are labeled with a single label. For example, when using (8,R) neighborhood, there are a total of 256 patterns, 58 of which are uniform, which yields in 59 different labels.
Ojala et al. (2002) noticed in their experiments with texture images that uniform patterns account for a little less than 90% of all patterns when using the (8,1) neighborhood and for around 70% in the (16,2) neighborhood. Each bin (LBP code) can be regarded as a micro-texton. Local primitives which are codified by these bins include different types of curved edges, spots, flat areas etc.
It should be easy to add this to the implementation, please open up a ticket with the specific feature request on http://code.opencv.org if you would like to see it implemented.
i lately tried implementing uniform LBP ( unfortunately not using opencv ).
it can be easily done using a lookup table
in the (8,R) case, it reduced the histogram size(and the size of the training dataset) to 22% of the original, and had no visible loss in recognition
.. might try again with opencv now ;)
Then make an answer out of it, so I can upvote at least!
@berak could you please share code for uniform LBP using look-up table? badly needed.....is it in C++ or OpenCV? thanks!
Asked: 2013-03-04 13:10:51 -0600
Seen: 552 times
Last updated: Mar 06 '13
yml will take much less space/time than xml
alternatively, you could retrieve the histograms after the training, like:
and do your own (binary) serialization