mirrors
/
research-ms-loss
mirror of https://github.com/msight-tech/research-ms-loss.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
MS-Loss: Multi-Similarity Loss for Deep Metric Learning
20 Commits
4 Branches
0 Tags
176 KiB
Python 99%
Shell 1%
 
 
Go to file
Matt Scott b68507d4e2
Merge pull request #25 from msight-tech/dependabot/pip/pillow-8.3.2 
Bump pillow from 6.2.0 to 8.3.2
 		2021-10-28 07:44:17 +08:00
configs Added ResNet50 to backbones 2019-10-21 12:55:36 +09:00
misc initial commit 2019-08-29 15:49:29 +08:00
ret_benchmark Adding XBM module 2020-03-08 14:54:30 -04:00
scripts Scripts for downloading CUB data-set and preparing train.txt/test.txt 2019-10-15 20:16:19 +09:00
tools initial commit 2019-08-29 15:49:29 +08:00
.flake8 initial commit 2019-08-29 15:49:29 +08:00
.gitignore initial commit 2019-08-29 15:49:29 +08:00
LICENSE initial commit 2019-08-29 15:49:29 +08:00
README.md Update README.md 2020-10-22 19:47:39 +08:00
ThirdPartyNotices.txt initial commit 2019-08-29 15:49:29 +08:00
requirements.txt Bump pillow from 6.2.0 to 8.3.2 2021-09-08 01:15:29 +00:00
setup.py initial commit 2019-08-29 15:49:29 +08:00

README.md

License: CC BY-NC 4.0

Multi-Similarity Loss for Deep Metric Learning (MS-Loss)

Code for the CVPR 2019 paper Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning

Performance compared with SOTA methods on CUB-200-2011

Rank@K 1 2 4 8 16 32
Clustering64 48.2 61.4 71.8 81.9 - -
ProxyNCA64 49.2 61.9 67.9 72.4 - -
Smart Mining64 49.8 62.3 74.1 83.3 -
Our MS-Loss64 57.4 69.8 80.0 87.8 93.2 96.4
HTL512 57.1 68.8 78.7 86.5 92.5 95.5
ABIER512 57.5 68.7 78.3 86.2 91.9 95.5
Our MS-Loss512 65.7 77.0 86.3 91.2 95.0 97.3

Prepare the data and the pretrained model

The following script will prepare the CUB dataset for training by downloading to the ./resource/datasets/ folder; which will then build the data list (train.txt test.txt):

./scripts/prepare_cub.sh

Download the imagenet pretrained model of bninception and put it in the folder: ~/.torch/models/.

Installation

pip install -r requirements.txt
python setup.py develop build

Train and Test on CUB200-2011 with MS-Loss

./scripts/run_cub.sh

Trained models will be saved in the ./output/ folder if using the default config.

Best recall@1 higher than 66 (65.7 in the paper).

Contact

For any questions, please feel free to reach

github@malongtech.com

Citation

If you use this method or this code in your research, please cite as:

@inproceedings{wang2019multi,
title={Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning},
author={Wang, Xun and Han, Xintong and Huang, Weilin and Dong, Dengke and Scott, Matthew R},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
pages={5022--5030},
year={2019}
}

License

MS-Loss is CC-BY-NC 4.0 licensed, as found in the LICENSE file. It is released for academic research / non-commercial use only. If you wish to use for commercial purposes, please contact sales@malongtech.com.