Proxy-Anchor-CVPR2020
Official PyTorch implementation of CVPR 2020 paper Proxy Anchor Loss for Deep Metric Learning.
A standard embedding network trained with Proxy-Anchor Loss achieves SOTA performance and most quickly converges.
This repository provides source code of experiments on four datasets (CUB-200-2011, Cars-196, Stanford Online Products and In-shop) and pretrained models.
Accuracy in [email protected] versus training time on the Cars-196
Requirements
- Python3
- PyTorch (> 1.0)
- NumPy
- tqdm
- wandb
- Pytorch-Metric-Learning
Datasets
-
Download four public benchmarks for deep metric learning
- CUB-200-2011
- Cars-196 (Img, Annotation)
- Stanford Online Products (Link)
- In-shop Clothes Retrieval (Link)
-
Extract the tgz or zip file into
./data/
(Exceptionally, for Cars-196, put the files in a./data/cars196
)
Training Embedding Network
Note that a sufficiently large batch size and good parameters resulted in better overall performance than that described in the paper. You can download the trained model through the hyperlink in the table.
CUB-200-2011
- Train a embedding network of Inception-BN (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model bn_inception \
--embedding-size 512 \
--batch-size 180 \
--lr 1e-4 \
--dataset cub \
--warm 1 \
--bn-freeze 1 \
--lr-decay-step 10
- Train a embedding network of ResNet-50 (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model resnet50 \
--embedding-size 512 \
--batch-size 120 \
--lr 1e-4 \
--dataset cub \
--warm 5 \
--bn-freeze 1 \
--lr-decay-step 5
Method | Backbone | [email protected] | [email protected] | [email protected] | [email protected] |
---|---|---|---|---|---|
Proxy-Anchor512 | Inception-BN | 69.1 | 78.9 | 86.1 | 91.2 |
Proxy-Anchor512 | ResNet-50 | 69.9 | 79.6 | 86.6 | 91.4 |
Cars-196
- Train a embedding network of Inception-BN (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model bn_inception \
--embedding-size 512 \
--batch-size 180 \
--lr 1e-4 \
--dataset cars \
--warm 1 \
--bn-freeze 1 \
--lr-decay-step 20
- Train a embedding network of ResNet-50 (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model resnet50 \
--embedding-size 512 \
--batch-size 120 \
--lr 1e-4 \
--dataset cars \
--warm 5 \
--bn-freeze 1 \
--lr-decay-step 10
Method | Backbone | [email protected] | [email protected] | [email protected] | [email protected] |
---|---|---|---|---|---|
Proxy-Anchor512 | Inception-BN | 86.4 | 91.9 | 95.0 | 97.0 |
Proxy-Anchor512 | ResNet-50 | 87.7 | 92.7 | 95.5 | 97.3 |
Stanford Online Products
- Train a embedding network of Inception-BN (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model bn_inception \
--embedding-size 512 \
--batch-size 180 \
--lr 6e-4 \
--dataset SOP \
--warm 1 \
--bn-freeze 0 \
--lr-decay-step 20 \
--lr-decay-gamma 0.25
Method | Backbone | [email protected] | [email protected] | [email protected] | [email protected] |
---|---|---|---|---|---|
Proxy-Anchor512 | Inception-BN | 79.2 | 90.7 | 96.2 | 98.6 |
In-Shop Clothes Retrieval
- Train a embedding network of Inception-BN (d=512) using Proxy-Anchor loss
python train.py --gpu-id 0 \
--loss Proxy_Anchor \
--model bn_inception \
--embedding-size 512 \
--batch-size 180 \
--lr 6e-4 \
--dataset Inshop \
--warm 1 \
--bn-freeze 0 \
--lr-decay-step 20 \
--lr-decay-gamma 0.25
Method | Backbone | [email protected] | [email protected] | [email protected] | [email protected] | [email protected] |
---|---|---|---|---|---|---|
Proxy-Anchor512 | Inception-BN | 91.9 | 98.1 | 98.7 | 99.0 | 99.1 |
Evaluating Image Retrieval
Follow the below steps to evaluate the provided pretrained model or your trained model.
Trained best model will be saved in the ./logs/folder_name
.
# The parameters should be changed according to the model to be evaluated.
python evaluate.py --gpu-id 0 \
--batch-size 120 \
--model bn_inception \
--embedding-size 512 \
--dataset cub \
--resume /set/your/model/path/best_model.pth
Acknowledgements
Our code is modified and adapted on these great repositories:
Other Implementations
- Pytorch, Tensorflow and Mxnet implementations (Thank you for Geonmo Gu :D)
Citation
If you use this method or this code in your research, please cite as:
@inproceedings{kim2020proxy,
title={Proxy Anchor Loss for Deep Metric Learning},
author={Kim, Sungyeon and Kim, Dongwon and Cho, Minsu and Kwak, Suha},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
year={2020}
}