RoMa
RoMa copied to clipboard
Parskatt
[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.
Hi, Johan, I have a question about the coordinates normalization in RoMa. https://github.com/Parskatt/RoMa/blob/36389eff4089d7b3f4a752c054e8fd0714af7177/roma/models/matcher.py#L565-L572 As you're using colmap coordinate fasion, the left- and right-most pixel should be `0.5` and `W-0.5`. The...
Thank you for your wonderful work. Could you please share your training logs? Another question is, if I want to train your network with a general dataset (the original training...
What is the overall reasoning speed? Are there relevant test data? tank you.
Hi, thanks for your outstanding work, I would like to ask if there is an inference code for windows operating system, and if configuring linux environment on wins system affects...
Hi Author, Thanks for sharing. Have you ever tried the performance of using tiny dino in the encoder part?
Hi, thanks for your great work! I noticed there's only an outdoor Tiny RoMa model available, is there (or is there going to be) an indoor tiny RoMa model available?
Dear authors, I notice there may be some disparities between Eq.(18) and implementation: 1. In robust_loss.py, Ln.92, the robust regression loss is implemented as: Compared to what is defined in...
Hi, thanks for this great repo. I was wondring from where i can change max number of features and matches that model will detect and try to match, similarly where...
