search

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
Apache License 2.0
901 stars 92 forks source link

Can we train XFeat on a custom dataset? #42

Open blackmamba-ops opened 1 month ago

guipotje commented 1 month ago

Hi @blackmamba-ops, yes, it is possible to train XFeat on a custom dataset. You just need to adapt the dataloader for your dataset.

Lannist commented 1 month ago

Hi @blackmamba-ops, yes, it is possible to train XFeat on a custom dataset. You just need to adapt the dataloader for your dataset.

How should I create this dataset, any Tutorial？

guipotje commented 4 weeks ago

All you need is to construct a data loader that delivers a correspondence map in the form xy_source (N,2) and xy_target (N,2), where each i-th line in the first set is the source pixel coordinates that corresponds to the i-th target pixel. This can be obtained via homography for synthetic training, or depth maps / dense optical flow. Here is the exact line where the code expects this correspondence map.