Issues
- 1
About the CE loss
#38 opened by xushangnjlh - 0
Why DinoV2 Large?
#37 opened by felipecadar - 0
How to use the DeDoD v2-L+RoMa ?
#36 opened by zhengshunkai - 0
setup problems
#35 opened by wuzelei123 - 5
question about detector and discriptor
#34 opened by ye450450 - 2
Possible bug ?
#33 opened by fratopa - 3
Unsure how to use the S-Version
#32 opened by geoeo - 12
- 2
DeDoDe doesn't produce absolute confidence?
#31 opened by nnop - 2
consulting
#29 opened by cuffsword - 1
- 1
What does the inv_temp variable mean?
#26 opened by hnyll - 2
Scores for keypoints?
#25 opened by hnyll - 3
about ‘prep_scene_info’
#24 opened by wangerniuniu - 5
about ‘prep_scene_info’
#23 opened by wangerniuniu - 2
the performance of dedode with superglue?
#22 opened by hkd724 - 20
- 2
How to train on my data
#20 opened by GavinYang5 - 4
"W" not in this function
#19 opened by gersonsolaris - 1
- 1
Training on own data set without groundtruths
#17 opened by Meda2997 - 1
Question about DeDoDe fine-tuning
#16 opened by haksorus - 5
hloc integration?
#11 opened by catid - 1
About the parameter in detector and match?
#15 opened by ghoshaw - 2
Update mega-1500 results
#13 opened by Parskatt - 0
Code cleanup
#14 opened by Parskatt - 6
Issue with samples
#10 opened by pmoulon - 35
- 3
getting 0 image matching correspondences for some image pair when giving whole video as consecutive image frames
#4 opened by Meda2997 - 6
Is there a way to finetune my own task?
#8 opened by 52THANOS - 7
- 3
this is a error ,please help me ,thank you!
#1 opened by LooperzZ