YingChen

I'm very sorry, and now our open source process has become very complicated and demanding. We need to complete the OETR open-source process first, and we need to go through...

The OETR architecture is not suitable for input pairs of images containing more than two images. The original intention of designing OETR was to find the co-visible area between image...

Thank you for your reply. I have tried to add OANet directly to the matching for pose estimation before, but the performance slightly decreased. I'm looking forward to your pose...

Hi, You could jump to IMU.cpp/747(void PreIntegrate), and judgment u1 is availability or not. Secondly you need to ensure timestamp is Synchronized() in main function.

@hankyul2 It is very likely that the gradient was calculated during validation, which resulted in out of memory. You can try torch.set_grad_enabled(False) in your validation process.