Issues
- 0
inference code for windows system
#40 opened by jlzhou1 - 1
- 0
- 0
- 3
- 4
- 1
Pairing Of Any Custom Points
#36 opened by cxb1998 - 9
Finding matches with arbitrary query points
#24 opened by justachetan - 4
- 7
Should KDE performs on the two images separately?
#34 opened by nnop - 4
Training code in `roma_indoor.py` seems to use non-distributed sampler with DDP
#32 opened by justachetan - 4
small model backbone
#28 opened by gc625-kodifly - 2
onnx
#15 opened by sushi31415926 - 0
about dataset
#27 opened by fighterzzzh - 1
Load indoor model?
#25 opened by RSKothari - 2
- 2
New commits give worse performance
#19 opened by Parskatt - 13
Training Plot
#18 opened by sushi31415926 - 13
Attempted to integrate into hloc
#8 opened by catid - 2
Is the indoor model licesned under MIT as well?
#17 opened by dawood95 - 1
VGG19
#16 opened by sushi31415926 - 2
doplegangers
#14 opened by scott-vsi - 3
Dino Version
#12 opened by sushi31415926 - 4
batch size
#13 opened by sushi31415926 - 4
what the delta_cls is used for?
#11 opened by lyp19 - 9
Sharing the checkpoint
#1 opened by Cuistiano - 2
- 4
Trying different backbone ?
#6 opened by tcourat - 2
About training settings
#7 opened by sugar-fly - 3
out of cuda memory
#5 opened by linqingfan - 14
- 8
Wrong version of pytorch (without gpu)
#4 opened by mpizenberg