issues
search
Parskatt
/
RoMa
[CVPR 2024] RoMa: Robust Dense Feature Matching; RoMa is the robust dense feature matcher capable of estimating pixel-dense warps and reliable certainties for almost any image pair.
https://parskatt.github.io/RoMa/
MIT License
430
stars
32
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
rename: roma -> romatch
#50
Vincentqyw
closed
1 day ago
1
About coordinates normalization
#49
nnop
opened
5 days ago
0
How does T_1to2 generated?
#48
lnexenl
closed
2 days ago
1
Running the demo is not working
#47
geodesit
opened
1 week ago
2
would you please share you log?
#46
smile1314-code
opened
1 week ago
0
Release some test dataset to evaluate
#45
skill-diver
opened
2 weeks ago
2
fix: line order in RobustLoss
#44
lnexenl
closed
2 weeks ago
2
About nan and inf grads
#43
lnexenl
closed
2 weeks ago
6
What is the overall reasoning speed? Are there relevant test data?
#42
tianyi-1349
opened
3 weeks ago
0
Tiny roma v1
#41
Parskatt
closed
3 weeks ago
0
inference code for windows system
#40
jlzhou1
opened
1 month ago
0
How to wrap image on another image based on current code base
#39
hi-zhangjie
opened
1 month ago
1
Failure to generate 3D reconstruction with unknown pose of the cameras
#38
royw99
opened
2 months ago
1
When would you release arbitrary keypoints matching?
#37
jimazeyu
opened
2 months ago
2
Pairing Of Any Custom Points
#36
cxb1998
opened
2 months ago
1
feat: handle pillow images as input
#35
dgcnz
closed
1 week ago
1
Should KDE performs on the two images separately?
#34
nnop
closed
2 months ago
7
Make it work with grayscale images
#33
Dawars
closed
2 months ago
1
Training code in `roma_indoor.py` seems to use non-distributed sampler with DDP
#32
justachetan
closed
2 months ago
4
feat: add option for mixed precision dtype to enable running on cpus
#31
dgcnz
closed
2 months ago
2
The Difference Between "forward_symmetric" and "forward"
#30
fighterzzzh
opened
2 months ago
4
Value of `coarse_res` and `upsample_res` for ScanNet
#29
justachetan
opened
2 months ago
3
small model backbone
#28
gc625-kodifly
opened
2 months ago
4
about dataset
#27
fighterzzzh
closed
2 months ago
0
RuntimeError: "LayerNormKernelImpl" not implemented for 'Half
#26
RSKothari
opened
3 months ago
4
Load indoor model?
#25
RSKothari
opened
3 months ago
1
Finding matches with arbitrary query points
#24
justachetan
opened
3 months ago
10
Update kde.py
#23
ducha-aiki
opened
4 months ago
4
approx kde
#22
Parskatt
closed
4 months ago
5
Recent update greatly increased GPU memory usage
#21
ducha-aiki
opened
4 months ago
2
Dev
#20
Parskatt
closed
5 months ago
0
New commits give worse performance
#19
Parskatt
closed
5 months ago
2
Training Plot
#18
sushi31415926
opened
5 months ago
13
Is the indoor model licesned under MIT as well?
#17
dawood95
closed
6 months ago
2
VGG19
#16
sushi31415926
closed
6 months ago
1
onnx
#15
sushi31415926
opened
6 months ago
2
doplegangers
#14
scott-vsi
closed
6 months ago
2
batch size
#13
sushi31415926
opened
7 months ago
4
Dino Version
#12
sushi31415926
opened
7 months ago
3
what the delta_cls is used for?
#11
lyp19
opened
7 months ago
4
Edit setup.py to be similar to setup.py of DKM repo
#10
qkqhd222
closed
5 months ago
1
License
#9
kbrodt
closed
9 months ago
2
Attempted to integrate into hloc
#8
catid
closed
5 months ago
13
About training settings
#7
sugar-fly
closed
10 months ago
2
Trying different backbone ?
#6
tcourat
opened
10 months ago
4
out of cuda memory
#5
linqingfan
closed
11 months ago
3
Wrong version of pytorch (without gpu)
#4
mpizenberg
closed
11 months ago
8
remove hardcoded .half casts and bf16
#3
Parskatt
closed
11 months ago
0
Unexpected floating ScalarType in at::autocast::prioritize
#2
urbste
closed
11 months ago
14
Sharing the checkpoint
#1
Cuistiano
closed
1 year ago
9
Next