Bag of Negatives for Siamese Architectures
Bojana Gajic (Computer Vision Center), Ariel Amato (Vintra, Inc.), Ramón Baldrich (Computer Vision Center), Carlo Gatta (Vintra, Inc.) AbstractTraining a Siamese architecture for re-identification with a large number of identities is a challenging task due to the difficulty of finding relevant negative samples efficiently. In this work we present Bag of Negatives (BoN), a method for accelerated and improved training of Siamese networks that scales well on datasets with a very large number of identities. BoN is an efficient and loss-independent method, able to select a bag of ``high quality negatives'', based on a novel online hashing strategy.
DOI
10.5244/C.33.180
https://dx.doi.org/10.5244/C.33.180
Files
Paper (PDF)BibTeX
@inproceedings{BMVC2019,
title={Bag of Negatives for Siamese Architectures},
author={Bojana Gajic and Ariel Amato and Ramón Baldrich and Carlo Gatta},
year={2019},
month={September},
pages={180.1--180.13},
articleno={180},
numpages={13},
booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
publisher={BMVA Press},
editor={Kirill Sidorov and Yulia Hicks},
doi={10.5244/C.33.180},
url={https://dx.doi.org/10.5244/C.33.180}
}
title={Bag of Negatives for Siamese Architectures},
author={Bojana Gajic and Ariel Amato and Ramón Baldrich and Carlo Gatta},
year={2019},
month={September},
pages={180.1--180.13},
articleno={180},
numpages={13},
booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
publisher={BMVA Press},
editor={Kirill Sidorov and Yulia Hicks},
doi={10.5244/C.33.180},
url={https://dx.doi.org/10.5244/C.33.180}
}