Eric Brachmann

Sorry to reply with a dad sentence, but: "Das Leben ist kein Ponyhof."

Unfortunately, many reviewers postpone any action until the last possible moment. So they join the discussion and insert their final rating at the same time, shortly before the deadline. 😔

I have seen more active discussions in the past. Maybe because this discussion period is short (<1 week)

It's time, dear #CVPR2026 reviewers. Time to decide.

If anything, I suspect that these tools will cover holes with persuasive bla bla. I do not think that authors are doing themselves a favour in the long run.

(sorry for quote posting, not allowed to answer...)

Interviews for our 2026 research internships will commence soon. It is not too late to apply! A lot of exciting stuff is happening in 3D vision, let's explore it together.

Often, those still refer to arXiv when the paper has been accepted to a conference or journal long ago 😕

If arXiv adds hallucination checks, it will be for sure exploited as an AI slob detection service before submitting to conferences...

Cmon. The things Dmytro mentions are essential building blocks that enable your research, done by people still living, whose career and funding(!) depends on that recognition. If we cite our software packages, datasets, pre-trained components and baselines, we are already well beyond 10, usually.

True. I think the CVPR forms are fine. And: the majority of reviews in my AC batch are solid or better.

As an AC, I try to prevent this pre-rebuttal and remind reviewers that rebuttal space is limited. But some people are irritatingly stubborn.

Agreed, but.

The reviewer: Here is an assortment of 20 things I do not like about the paper, in random order, things big and small. Suggestions for the rebuttal: See weaknesses. Post rebuttal: The authors addressed items 1-19 but not item 20. Reject.

AC pro tip: When reading reviews, mentally substitute the "confidence" rating (low, medium, high) with "overconfidence", and everything is much better calibrated.

They made it!!! *happy AC*

Ah, but I want to believe we are just very good at reviewer bidding?

Your point might be more plausible, though. At least for authors - and I guess this group is big enough...

a man in a blue shirt is sitting at a desk with a pepsi sign on the wall ALT: a man in a blue shirt is sitting at a desk with a pepsi sign on the wall

Reviewers certainly had a very productive night. Went from 12 missing to 1 missing. Routing for the last one: You can do it!

Would be nice if OR implemented AB testing, so we knew the impact of these choices. Sample sizes are big enough these days I wager.

I would never say no to that 😉

Let's put this checkbox inside OR, after registration, so you need to do a few clicks to reach it. Then, we have a pre-deadline by which you have to click it. If you click it in time, you have proven that you can adhere to deadlines and hence are spared the reminders. Dips on "the Brachmann check".

Depends on the scope and variation of t. 3D is unbounded which is a real pain. And then the joy of SO3 🫣

And if we are very lucky, we have something halfway reasonable for 3D by 2036.

Ignacio does not seem to be on here... but nice website :D
ialzugaray.github.io

What a nice morning to find that Alzugaray, Taher @marwantaher.bsky.social , Davison @ajdavison.bsky.social have turned ACE into a SLAM system. Since we did ACE, I was wondering whether it was possible, and I always thought who if not those folks can pull it off. 🙂

ACESLAM: arxiv.org/abs/2512.14032

You know what season it is! Right. Internship application season.

Niantic Spatial is offering research internships on a multitude of 3D vision topics: relocalization, reconstruction, 3D VLMs... Top tier papers regularly come out of our internships 🚀

nianticspatial.careers.hibob.com/jobs/0fc4871...

I like it. I hear it in captain Picards voice.

What fresh hell is this?

Is that not like saying "People will not submit AI generated papers, because only good papers get in"?

So, in the spirit of the original message, we would have to charge for reviewing, and good reviewers will get their investment back with interest.

A skill I have developed through years of looking for individual pieces in huge piles of LEGO.

Robotics researchers, this is an occasion not to be missed: top-notch speakers in an excellent location in the Alps this February 🤖⛷️
It's the Winter School on Social Robotics, Artificial Intelligence and Multimedia (SoRAIM), 9-13 Feb 2026 👇

That picture is an interesting way to point out that @chriswolfvision.bsky.social has no shared letter between his first and last name...

Oh, wow.