It's been several weeks since I started my research internship at @nianticspatial.bsky.social in London. Excited to finally work with @ericbrachmann.bsky.social and the team!
Eric Brachmann
@ericbrachmann.bsky.social
Niantic Spatial, Research. Throws machine learning at traditional computer vision pipelines to see what sticks. Differentiates the non-differentiable. ๐Europe ๐ http://ebrach.github.io
ACTUALLY, that very same asteroid was already discovered 1990 by the hobby astronomer, Jรผrgen S.
I think I am ready to be replaced by AI. My whole mentoring can be summarized by the following expert system: if True: print("Visualize everything.")
I have had 70 co-authors so far. I took some time to wire all these names up with their websites on my personal page. (The BOP papers have been a real pain. ๐ ) Fun to browse through what all these people have been up to since we collaborated.
We are grateful to all of the 17,491 reviewers who helped make #CVPR2026 possible. We are especially pleased to recognize the following Outstanding Reviewers, whose high-quality reviews (as judged by their Area Chairs) placed them among the top 5% of reviewers. 1/2
In my team we have two open positions on AI for robotics and in particular on manipulation, for junior and senior researchers. With us, connect fundamental research questions with real operational problems and contribute to emerging AI-driven robotics services. careers.werecruit.io/en/naver-lab...
It seems, that we have failed the communication about IMC26. Let's try again. The competition this year is here: kaggle.com/competitions... No prizes, but whole year leaderboard -- similar to KITTY and other academic competitions. 3D people, please retweet and share.
Image Matching Challenge 2025 Ongoing
Ongoing leaderboard for Image Matching Challenge 2025.
kaggle.com
We were pretty mad at the reviewer but they turned out to be reasonable. I understand the frustration that classical methods are often overlooked. Here, it was a genuine oversight. I did not know the E5+1 solver. Including it made the paper stronger. Full discussion: openreview.net/forum?id=rmD...
A Scene is Worth a Thousand Features: Feed-Forward Camera...
Visually localizing an image, i.e., estimating its camera pose, requires building a scene representation that serves as a visual map. The representation we choose has direct consequences towards...
openreview.net
David Nordstrรถm@davnords.bsky.social ยท 4mo agoSide note: pretty cool you managed to get a reviewer to update from 2 to 8 after a strong rebuttal. Will live on that hopium for ECCV rebuttals...
FastForward has been accepted to #ICLR2026 and will be presented there next week. If you go, say hi to @axelbarroso.bsky.social. If not, don't fret. Axel has prepared a short video for you: www.youtube.com/watch?v=r0Qk...
[ICLR 2026] Feed-Forward Camera Localization from a Collection of Image Features
YouTube video by Niantic Spatial
youtube.com
Eric Brachmann@ericbrachmann.bsky.social ยท 10mo agoโฉ FastForward โฉ A new model for efficient visual relocalization. Accurate camera poses without building structured 3D maps. nianticspatial.github.io/fastforward/ Work by @axelbarroso.bsky.social, Tommaso Cavallari and Victor Adrian Prisacariu.
๐ค
Vision and Graphics Trends@si-cv-graphics.bsky.social ยท 4mo ago๐๐ป๐ ๐ฏ๐ ๐ฆ๐ฐ๐ฒ๐ป๐ฒ ๐ถ๐ ๐ช๐ผ๐ฟ๐๐ต ๐ญ๐ ๐ง๐ผ๐ธ๐ฒ๐ป๐: ๐ฏ๐-๐๐ฟ๐ผ๐๐ป๐ฑ๐ฒ๐ฑ ๐ฅ๐ฒ๐ฝ๐ฟ๐ฒ๐๐ฒ๐ป๐๐ฎ๐๐ถ๐ผ๐ป ๐ณ๐ผ๐ฟ ๐ฆ๐ฐ๐ฒ๐ป๐ฒ ๐๐ฒ๐ป๐ฒ๐ฟ๐ฎ๐๐ถ๐ผ๐ป ๐ฎ๐ ๐ฆ๐ฐ๐ฎ๐น๐ฒ Dongxu Wei, Qi Xu, Zhiqi Li ... Peidong Liu arxiv.org/abs/2604.11331 Trending on www.scholar-inbox.com
Thanks, but no thanks. You might get me after all if we submit, but for now it feels good to decline. Last year, it was a horrible experience to review for NeurIPS. And I had a reduced load of two papers....
Eric Brachmann@ericbrachmann.bsky.social ยท 11mo agoInteresting, you are right, the general process did not change in the last few years. I did not review for NeurIPS in a long time. I volunteered this year but would not do it again in the current setting. It is an unreasonable amount of work compared to eg *CCV/CVPR. And ACing must be a nightmare.
Even ScholarInbox wants me to focus. "Today, we found 1 (of 371) articles relevant for you. Relevance: -17"
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...)
TThomas Dietterich@tdietterich.bsky.social ยท 6mo ago
For me, writing is the process of assembling a logical narrative, and it is during the writing process that I discover the holes in my reasoning. I'm very skeptical that current AI tools are capable of doing this.
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.
Eric Brachmann@ericbrachmann.bsky.social ยท 8mo agoYou 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 ๐ nianticspatial.careers.hibob.com/jobs/0fc4871...
AC pro tip: When reading reviews, mentally substitute the "confidence" rating (low, medium, high) with "overconfidence", and everything is much better calibrated.
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. ๐ 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 ๐ nianticspatial.careers.hibob.com/jobs/0fc4871...
Careers
nianticspatial.careers.hibob.com
Robotics researchers, this is an occasion not to be missed: top-notch speakers in an excellent location in the Alps this February ๐คโท๏ธ It's the Winter School on Social Robotics, Artificial Intelligence and Multimedia (SoRAIM), 9-13 Feb 2026 ๐
Recording of the 10th R6D workshop #ICCV2025 w/ results of the 2025 BOP challenge is out: youtu.be/NymS-f4DNh0 3.5h of goodness dedicated to robot-object manipulation w/ Hao-Shu Fang (MIT), Maximilian Durner (DLR), Agastya Kalra+Vahe Taamazyan (Intrinsic) and Sergey Levine (Physical Intelligence)
2738 Final 10th International Workshop on Recovering 6D Object Pose
YouTube video by ComputerVisionFoundation Videos
youtu.be
Of all my multiple Lufthansa flights on this journey, the only one that was punctual was the one with a short connection time, and that I therefore missed due to the previous one being late. Now, after one involuntary overnight stay: delay again. Great job @lufthansagroup.bsky.social
Coarse plan for a 1-week rebuttal period: day 1: read the reviews day 2-6: calm down day 7: write the rebuttal
โผ๏ธ Major update to the #ACEZero code base. We added the capabilities of the #ICCV2025 paper on Scene Coordinate Reconstruction Priors. Gives you, e.g., a RGB-D version of ACE0. In the snipped below, I jointly reconstruct multiple RGB-D scans of ScanNet. github.com/nianticlabs/...
Eric Brachmann@ericbrachmann.bsky.social ยท 10mo ago๐ Scene Coordinate Reconstruction Priors at #ICCV2025 ๐ Can we learn what a successful reconstruction looks like, and use this knowledge when reconstructing new scenes? Explainer: youtu.be/RkV6U5xYc20 Project Page: nianticspatial.github.io/scr-priors Inspiring work by Wenjing Bian et al.!
Code is out. We provide the pre-trained ACE-G model, as well as code to replicate the paper results, and to evaluate ACE-G on new scenes. #ICCV2025 github.com/nianticspati...
GitHub - nianticspatial/ace-g: [ICCV 2025] ACE-G is an architecture and pre-training scheme to improve generalization for scene coordinate regression-based visual relocalization.
[ICCV 2025] ACE-G is an architecture and pre-training scheme to improve generalization for scene coordinate regression-based visual relocalization. - nianticspatial/ace-g
github.com
Eric Brachmann@ericbrachmann.bsky.social ยท 10mo ago๐ฅ ACE-G, the next evolutionary step of ACE at #ICCV2025 ๐ฅ We disentangle coordinate regression and latent map representation which lets us pre-train the regressor to generalize from mapping data to difficult query images. Page: nianticspatial.github.io/ace-g/ Stellar work by Leonard Bruns et al.!
Hi everyone. Can we please stop naming methods in leetspeak? It slows down my writing - it is a skill I happily abandoned 20 years ago.
10th (!!) R6D workshop @ ICCV 2025: cmp.felk.cvut.cz/sixd/worksho... ๐ค Object pose estimation for industrial robotics ๐ข Speakers: - Sergey Levine (UC Berkeley) - Haoshu Fang (MIT) - Maximilian Durner (DLR) - Agastya Kalra & Vahe Taamazyan (Intrinsic) ๐ BOP 2025 with new BOP-Industrial datasets
After next week, I REALLY need a break from PowerPoint.
Eric Brachmann@ericbrachmann.bsky.social ยท 10mo agoLooking forward to a busy #ICCV2025. I will give three (very different) talks at workshops and tutorials, see info below. We also present two papers, ACE-G and SCR Priors. And it's the 10th (!) anniversary of the R6D workshop, which we co-organize.
