tjddus9597 Proxy-Anchor-CVPR2020 issues

tjddus9597 / Proxy-Anchor-CVPR2020

Official PyTorch Implementation of Proxy Anchor Loss for Deep Metric Learning, CVPR 2020

MIT License

312 stars 60 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Unfair comparison, your method has 100x higher lr

#28 grypesc opened 10 months ago
0
Evaluate SOP using pretrained model, R@1 much lower from the number on README.md

#27 junwei-h opened 1 year ago
0
BalancedSampler is highly inefficient and way too slow for even dataset with mild size

#26 jdhao closed 2 years ago
1
[bug] proxyNCA loss also has parameters that needs to optimized

#25 jdhao closed 2 years ago
3
fix redundant path join

#24 jdhao closed 2 years ago
1
model.model.embedding bug

#23 jjkim0807 closed 2 years ago
1
Error when I use "adamw" as the optimizer (UnboundLocalError: local variable 'beta1' referenced before assignment)

#22 farshadsm closed 3 years ago
2
Evaluation on ReID datasets

#21 itsnamgyu closed 3 years ago
1
Inference Get Class Prediction

#20 oggyfaker closed 3 years ago
1
N-Pair Loss reproducability

#19 alpoler closed 3 years ago
12
Nice topic

#18 mrk1992 closed 3 years ago
0
Using gt box or expanding gt box?

#17 FightStone closed 3 years ago
3
[UPDATE to code/train.py ] criterion = losses.Proxy_NCA().cuda()

#16 moured closed 3 years ago
1
CUDA out of memory

#15 ShuteLee closed 3 years ago
11
CUDA out of memory

#14 ShuteLee opened 3 years ago
3
Reproduce the performance in CUB

#13 liuhl-source closed 3 years ago
5
bug while evaluation with cub200

#12 sycophant-stone closed 3 years ago
0
Hyperparameters of Proxy_NCA

#11 say2sarwar closed 3 years ago
1
The hyperparameters of other methods

#10 mojie4321 closed 3 years ago
1
How did you leverage data-to-proxy relationship?

#9 jlian2 closed 3 years ago
1
Regarding training complexity

#8 jlian2 closed 3 years ago
2
Reproducing the results with trained weights

#7 birdortyedi closed 4 years ago
1
Experimental results using other methods

#6 kdhht2334 closed 4 years ago
2
Keras implementation of the loss function

#5 nixingyang closed 4 years ago
1
Hard Sample mining

#4 jlian2 closed 4 years ago
2
Distance measure

#3 jlian2 closed 4 years ago
1
random sampling

#2 EulerLee closed 3 years ago
6
backbone

#1 Bigwode closed 4 years ago
6