Note: the content of this page might change slightly; be sure to refresh right before you start
The goal of this assignment is to classify the images based on the feature vectors produced in the previous step.
This is step 3 of the pipeline shown in the image to the right.
You can use the complete set of training-images to train your machine-learning algorithm. It is good practice to do K-fold evaluation to test your algorithm when developing.
You will produce a script or program called classify. The
arguments that will be passed to this program are featurefile.feat
class.rec. The first argument is an existing feature file (containing a
single feature vector),
which must be read. The second argument is the file which should be written in
the following format:
label1 0.93
label2 0.38
label3 0.21
...
where the first column contains the class-label and the second column contains a (pseudo-)probability, i.e., how likely this image actually has this label.
Hand in your code in a .tar.gz file, marked with your name and assignment
number. If there are compilation steps, create a Makefile, which compiles
your program with a single make command.
Your code should have a basis in the literature. Put comments in your code detailing the steps you took and on which articles you base your method.
Last modified: April 27, 2011, by Jean-Paul van Oosten
Part of the HWR course