Eric Brachmann

Just from todays feed: bsky.app/profile/kmyi...

I think it is a great time to have such a tutorial again. As we see competitive RANSAC-free approaches arise, it is worth looking back - and looking forward.

For those going to @iccv.bsky.social, welcome to our RANSAC tutorial on October 2025 with
- Daniel Barath
- @ericbrachmann.bsky.social
- Viktor Larsson
- Jiri Matas
- and me
danini.github.io/ransac-2025-...
#ICCV2025

Whenever I do a voice over, I realize that I sound like Arnold Schwarzenegger when I say "coordinate". Unfortunately, in my line of work, I have to say "coordinate" a lot...

I thought you just turned 34?

Trendy.

Great to have Kwang on Bluesky! (And quite flattering that this is his first post 😳)

The model estimates image poses relative to images of which you already know the poses. It solves a sub-problem of what COLMAP generally is used for. FastForward is a visual relocalization model, not a full structure-from-motion model.

*whisper* Only one way to know...

Oh no, today of all days...

A Scene is Worth a Thousand Features: Feed-Forward Camera Localization from a Collection of Image Features

@axelbarroso.bsky.social, Tommaso Cavallari, Victor Adrian Prisacariu, @ericbrachmann.bsky.social

arxiv.org/abs/2510.00978

FastForward was trained on challenging frame sets. Therefore, it is able to solve difficult relocalization cases, like opposing shots in the video below. We only have mapping images of the front of the sign. Still FF relocalizes well even behind the sign.

FastForward uses 20 retrieved images by default. But no need to compute attention over 21 images. To keep the model lean, we sub-sample the feature tokens of mapping images.

Each token is combined with its ray embedding. Hence, FastForward operates on samples of visual tokens in scene space.

Building visual maps via reconstruction (SfM, SCR) can take time. Instead:

⏩ Get poses for your mapping images, e.g. via SLAM in real time.
⏩ Build a retrieval index in seconds.
⏩ FastForward predicts the relative pose between a query image and all retrieved images in one forward pass.

⏩ FastForward ⏩ 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.

I agree, could have been the sächsische Schweiz easily from the image.

"jokes and nonsense"

Add awkward motivational spins and you have LinkedIn.

Limitations of pseudo ground truth.

Our software engineering prof pretty much told us exactly that. And that you will earn a lot of money if you master that skill.

Small but important quality of life update by @rerun.io: Customizable frustum colors. Finally we can distinguish estimates and ground truth in the same 3D view ;)

Here we go again 😅 This time I’m planning to take a more senior role to help mentor the next gen of publicity chairs. Please consider volunteering!

Recovering 6D Object Pose - Looking forward to the R6D workshop at #ICCV2025 in Hawaii! (not looking forward to the jet lag)

cmp.felk.cvut.cz/sixd/worksho...

The accompanying challenge is still open for submissions until Oct 1st. And the workshop also still accepts non-proceeding paper submission!

The television:

"Nu."

Maybe it deals better with a French accent than with a Saxon accent 🫣

fry from futurama is shown with his eyes closed ALT: fry from futurama is shown with his eyes closed

not sure if...

That is probably correct.

I always found it to work out of the box for our CNN and MLP architectures... But then again, I never tried to tune it. Maybe I should have...

Many jump scares?

That guy. (remixed from @xkcd.com)