Code for Ragav Venkatesan, Parag Shridhar Chandakkar, Baoxin Li "Simpler non-parametric methods provide as good or better results to multiple-instance learning." at the IEEE International Conference on Computer Vision (ICCV), Santiago, Chile, 2015.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
src
.gitignore
License.md
README.md

README.md

This is the code for the paper

Ragav Venkatesan, Parag Shridhar Chandakkar, Baoxin Li "Simpler non-parametric methods provide as good or better results to multiple-instance learning." at the IEEE International Conference on Computer Vision (ICCV), Santiago, Chile, 2015.

The code also contains the DR dataset and one Synthetic datasets. To run the code, simply setup the parameters in MILLooper.m and run. The parameters are self-explanatory and are commented upon in a detailed manner in the code itself. The code only has the brute force optimization. Due to license issues I am not able to provide the EMDD version of the code, for which I used the original Y.Chen's executable. It is not difficult to implement from there on. The results are close but are not the same as reported because of the change in optimization methodology.

Thanks for using the code, hope you had fun.

Ragav Venkatesan http://www.public.asu.edu/~rvenka10