This is the code of max-margin multi-instance dictionary learning for image classification. 

Required libraries:
1. VLFeat, by Andrea Vedaldi et al. Please download it from: http://www.vlfeat.org/index.html
2. LibLinear, by Machine Learning Group at National Taiwan University. We modified the library for training using less memory. Our modified version is included in this package.

The code is written in Matlab and tested on 64bit Windows. To run this code:
1. Please download the Scene-15 dataset from: http://www.cs.illinois.edu/homes/slazebni/research/scene_categories.zip
2. Setup dataset path in "exp_sceneft.m" and run "exp_sceneft.m"

This code is free for using academic research. If you use this code, please cite the following paper:
Xinggang Wang, Baoyuan Wang, Xiang Bai, Wenyu Liu, and Zhuowen Tu, Max-Margin Multiple Instance Dictionary Learning, International Conference on Machine Learning (ICML), Atlanta, June, 2013.

For any question, please contact: Xinggang Wang (xgwang@hust.edu.cn)