This repository provides the implementation of NAMVH as described in the papers:
Nonlinear Asymmetric Multi-Valued Hashing
Cheng Da, Gaofeng Meng, Shiming Xiang, Kun Ding, Shibiao Xu, Qing Yang, and Chunhong Pan
TPAMI, 2018.
Meanwhile, this repository also provides the implementation of AMVH as described in the papers:
AMVH: Asymmetric Multi-Valued Hashing
Cheng Da, Shibiao Xu, Kun Ding, Gaofeng Meng, Shiming Xiang, Chunhong Pan
CVPR, 2017.
Linux 16.04.5 NVIDIA GPU + CUDA-8.0 and corresponding CuDNN
Caffe
Matlab 2015b
You need to compile the modified Caffe library in this repository. You can consult the generic Caffe installation guide.
When the dependencies of Caffe are installed, try:
make
If there are no error messages, then you can compile and install the matlab wrappers:
make matcaffe
run AMVH for AMVH: Asymmetric Multi-Valued Hashing on ESP-GAME
cd NAMVH
matlab -r -nodesktop run_AMVH on ESP-GAME with 8-bit codes.
maps =
0.4362 0.4906
0.4953 0.5190
noted: maps(1,1) = AMVH_bin; maps(1,2) = AMVH_real; maps(2,2) = AMVH_mul
run NAMVH for Nonlinear Asymmetric Multi-Valued Hashing on ESP-GAME with 8-bit codes.
cd NAMVH
matlab -r -nodesktop run_NAMVH
maps =
0.4822 0.5045
0.5208 0.5599
noted: maps(1,1) = NAMVH_1^b; maps(1,2) = NAMVH_1^r; maps(2,2) = NAMVH_{10}^r
We provide ESP-GAME dataset ./data to test our codes.
If you find NAMVH and AMVH useful in your research, please consider citing the following paper:
@ARTICLE{8451950,
author={C. Da and G. Meng and S. Xiang and K. Ding and S. Xu and Q. Yang and C. Pan},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={Nonlinear Asymmetric Multi-Valued Hashing},
year={2018},
volume={PP},
number={1},
pages={1-1},
doi={10.1109/TPAMI.2018.2867866},
ISSN={0162-8828},
}
@inproceedings{DBLP:conf/cvpr/DaXDMXP17,
author = {Cheng Da and
Shibiao Xu and
Kun Ding and
Gaofeng Meng and
Shiming Xiang and
Chunhong Pan},
title = {{AMVH:} Asymmetric Multi-Valued hashing},
booktitle = {{IEEE} Conference on Computer Vision and Pattern Recognition,{CVPR}},
pages = {898--906},
year = {2017},
url = {https://doi.org/10.1109/CVPR.2017.102},
doi = {10.1109/CVPR.2017.102},
}