Johan Edstedt

If only there was a benchmark bigger than 37 pairs... Alas.

Coming soon

Et tu3r

Putting a vowel before 3R should be illegal.

Want to mention that I don't think schedules are always bad... It does tend to be the case that lowering learning rate puts you in regions with higher gradients., which can lead to spikes. On the other hand, higher lr means more instability.

Edstedtian Representation Hypothesis.

You're on the exponential curve.

Nope, more scenes.

Other than being lefthanded it makes sense ;^)

This is just big wheat propaganda, flour cant scale.

Take me down to the Parallax city where the far moves slow and the near moves quickly

"a lazy human would not ignore all prior art" idk, the papers I've reviewed say otherwise ;)

I might try some approaches this weekend.

My guess is that:

1. COLMAP would work pretty well (with good corresp), but baseline is rather small, and dynamics of cloud would be tricky.

2. Feed-forward methods would underestimate the size of the cloud.

Interested if this is indeed the case, or my intuitions are wrong.

Is there any current 3/4D reconstruction method able to accurately reconstruct this scene? (with correct size of thunderstorm)

regardless of your views on AI, i strongly agree with this viewpoint! you are not funded by taxpayers to optimize your knowledge consumption workflows. it's good that you enjoy your job, you are getting paid to enjoy your job

Should the rebuttal contain the updated paper ID?

For revising after the OpenReview debacle we were not supposed to do so.

Sounds like it was "on time" by sj standards.

You have days like these, but also days where it looks like Mordor.

hehe, we'll see what I have time for.

Sure, but the DeDoDe descriptor would perform very well with 4k keypoints for another detector, such as ALIKED or DaD :)
I would bet that we would beat their IMC22 numbers in that setting.

If you run the dedode descriptor with dad keypoints you generally get much better results for lower number of keypoints than the original detector. (Shown is dedode descriptions matched with different detectors).

Yes, likely. We ran with 30k keypoints (or whatever gave the best performance). You can see that our results for other descriptors is also in general better.