oravus / lostx Goto Github PK

Hi,

Could you please generate the script of how to find the correspondences between the two matching images using the found descriptors? Not exactly sure on how to do that using the given format of the descriptors? Or atleast could you provide a brief method on how you did it?

Thanks.

Hello,

I'm hoping to implement sequence matching with LoST-X as presented in the paper, but I don't fully understand this statement in the paper:

We use a slight variant of the publicly available OpenSeqSLAM [26] with an induced ability to accumulate sequence scores for only the top matching candidates generated using the LoSTX framework

Can you clarify how this is performed or share the modification of OpenSeqSLAM that was used?

Thank you!

Thank you for your great work! Could you please provide me with your evaluation part of the code? Because i can't get the same result in the paper.

would you please upload trained model on SYNTHIA?