verlab / accelerated_features

Implementation of XFeat (CVPR 2024). Do you need robust and fast local feature extraction? You are in the right place!

https://www.verlab.dcc.ufmg.br/descriptors/xfeat_cvpr24

verlab/accelerated_features Issues

Minor typo in Section 3.1 of the paper
Updated 7 days ago1
How is the warping performed to generate the synthetic training set?
Updated 7 days ago1
OpenVINO support?
Closed 7 days ago3
Image registration based on xfeat c++
Updated 11 days ago
Image matching similarity
Updated 11 days ago1
Do you plan to convert this code into C++code with onnx? I think that's a good idea
Updated 12 days ago7
How does this method perform if training without Reliability Map?
Closed 12 days ago2
When the training code will be release？
Updated 13 days ago2
How does this method compare with LightGlue?
Closed 13 days ago3
Whether the image preprocessing needs to be divided by 255?
Closed 13 days ago2
Multi-modality?
Closed 13 days ago2
Will your training code contain training using homographies?
Closed 15 days ago3
how to use cv.BFMatcher() for matching
Closed 17 days ago1
PR to kornia?
Updated 23 days ago3
What is the expected input values?
Closed 24 days ago7