mihaidusmanu/d2-net
D2-Net: A Trainable CNN for Joint Description and Detection of Local Features
Jupyter NotebookNOASSERTION
Issues
- 4
SfM 3D reconstruction dataset
#55 opened - 3
Any suggestion for parallelization
#54 opened - 1
Multiscale running error
#53 opened - 4
InLoc dataset
#52 opened - 1
- 2
trained model src
#50 opened - 2
- 1
- 1
- 2
Smaller Subset Data Available?
#46 opened - 2
Is depth map really necessary?
#45 opened - 2
- 2
- 2
- 5
Using MobileNet instead of VGG16
#41 opened - 3
How to understand the loss function???
#40 opened - 0
How to achieve matching acceleration
#39 opened - 5
Download problem(Undistorted_sfm)
#38 opened - 7
- 3
Question about the training convergence
#36 opened - 7
Keypoint Scores
#35 opened - 2
Test Network Arch.
#34 opened - 1
- 1
- 1
- 2
Question about the training result
#30 opened - 2
- 8
- 5
D2-Net feature matching
#27 opened - 2
Preprocessed data
#26 opened - 6
Use of VGG16 pretrained weights
#25 opened - 2
- 4
About the localization benchmark
#23 opened - 3
Problem extracting multi-scale features
#22 opened - 2
Performance of trained model weights
#21 opened - 3
- 4
About Hpatches
#19 opened - 3
MegaDepth
#18 opened - 2
Data download error
#17 opened - 2
Whether to use ReLU for the last layer
#16 opened - 2
Overlap matrix resulting in many -1.
#15 opened - 2
Question about preparing the training data.
#14 opened - 2
- 1
Tensorflow version
#12 opened - 6
Unstable training on subset of megadepth
#11 opened - 6
Getting "scene_info" file.
#10 opened - 5
Enabling Visual Output
#9 opened - 8
Inliers for Less textured Images
#8 opened - 4
Invalid Models Links
#7 opened - 3