Sparse Contextual Activation for Efficient Visual Re-ranking

The Matlab implementation of Sparse Contextual Activation (SCA) published in IEEE Transactions on Image Processing (TIP) 2016. SCA is a very efficient and effective re-ranking algorithm for unsupervised object retrieval. Strong evidence also demonstrates that SCA can be easily applied to person re-identification.

Usage

To use the code, simply replace "d" with your own input distance matrix, and specify a proper value for k1 and k2.

Citation

Please cite this paper if it helps your research:

@article{SCA,
  title={Sparse contextual activation for efficient visual re-ranking},
  author={Bai, Song and Bai, Xiang},
  journal={IEEE Transactions on Image Processing},
  volume={25},
  number={3},
  pages={1056--1069},
  year={2016},
}