Readme for the SCDA method package version Dec. 18, 2016

The package includes the MATLAB code of the SCDA method for fine-grained image retrieval.

The main / key parts of SCDA are included in this package, i.e., selection and aggregation.

You will find a demo 'imdb.mat' file for the CUB200 dataset. But you need to download the pre-trained VGG-16 model by yourself via http://www.vlfeat.org/matconvnet/pretrained/.

References: X.-S. Wei, J.-H. Luo, J. Wu and Z.-H. Zhou. Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval. IEEE Transactions on Image Processing, 2016, in press.

ATTN: This packages are free for academic usage. You can run them at your own risk. For other purposes, please contact Prof. Jianxin Wu (wujx2001@gmail.com).

Requirement: To use this package, MatConvNet (http://www.vlfeat.org/matconvnet/) must be available.

ATTN2: This packages were developed by Mr. Xiu-Shen Wei (weixs.gm@gmail.com). For any problem concerning the code, please feel free to contact Mr. Wei.

---