Pytorch Implementation of Deep Orthogonal Fusion of Local and Global Features (DOLG)
This is the unofficial PyTorch Implementation of "DOLG: Single-Stage Image Retrieval with Deep Orthogonal Fusion of Local and Global Features"
reference: https://arxiv.org/pdf/2108.02927.pdf
Prerequisites
- PyTorch
- PyTorch Lightning
- timm
- sklearn
- pandas
- jpeg4py
- albumentations
- python3
- CUDA
Data
You can get the GLDv2 dataset from here.
If you just want the GLDv2-clean dataset, check this kaggle competition dataset.
Place your data like the structure below
data
├── train_clean.csv
└── train
└── ###
└── ###
└── ###
└── ###.jpg
Citations
@misc{yang2021dolg,
title={DOLG: Single-Stage Image Retrieval with Deep Orthogonal Fusion of Local and Global Features},
author={Min Yang and Dongliang He and Miao Fan and Baorong Shi and Xuetong Xue and Fu Li and Errui Ding and Jizhou Huang},
year={2021},
eprint={2108.02927},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
406 Dec 23, 2022
17 Sep 23, 2021
883 Jan 7, 2023
91 Dec 2, 2022
348 Dec 23, 2022
2 Oct 13, 2021
873 Jan 6, 2023
123 Dec 28, 2022
5 Jul 21, 2022
184 Dec 12, 2022
502 Jan 3, 2023
104 Nov 29, 2022
2.6k Jan 2, 2023
11 Aug 28, 2022
906 Jan 3, 2023
126 Dec 6, 2022