Valentina Zadrija

Nighttime cityscape of Tokyo with dense illuminated skyscrapers and Tokyo Tower glowing orange on the right. Centered over the skyline is the EU–Japan logo combining the European Union and Japanese flags, with the text “EU–JAPAN” and “HORIZON EUROPE” beneath it. The European Commission logo appears in the bottom-right corner.

Japan is joining Horizon Europe, the EU’s flagship research and innovation programme.

Openness and international cooperation can shape a bright future for science and technology.

Through science, we can build bridges, strengthen competitiveness, and accelerate the green and digital transitions.

On Geometric Understanding and Learned Data Priors in VGGT
Jelena Bratulić, Sudhanshu Mittal, Thomas Brox, Christian Rupprecht
tl;dr: all magic happens in the middle layers (as tested on the ShapeNet)
arxiv.org/abs/2512.11508

Attending #Neurips2025? Get your personalized Scholar Inbox conference program now to easily navigate the poster sessions and find what you are looking for:
www.scholar-inbox.com/conference/n...

A student with a laptop can change the world 🚀
Thank you for the live streaming - the talks were great 🙏

The Principles of Diffusion Models This monograph presents the core principles that have guided the development of diffusion models, tracing their origins and showing how diverse formulations arise from shared mathematical ideas. Diffu...

"The Principles of Diffusion Models" by Chieh-Hsin Lai, Yang Song, Dongjun Kim, Yuki Mitsufuji, Stefano Ermon. arxiv.org/abs/2510.21890
It might not be the easiest intro to diffusion models, but this monograph is an amazing deep dive into the math behind them and all the nuances

Kickstarting our workshop on Flow matching and Diffusion with a talk by Eric Vanden Eijnden on how to optimize learning and sampling in Stochastic Interpolants!

Broadcast available at gdr-iasis.cnrs.fr/reunions/mod...

Hirose et al. OmniVLA: An Omni-Modal Vision-Language-Action Model for Robot Navigation
arxiv.org/pdf/2509.19480

Navigation VLA (as opposed to robotic manipulation) + conditioning on language, ego-centric & satellite images, GPS-ish poses
2 flavours: 7B Llama + 50M edge version (EfficientNet, CLIP)

Major announcement ✨registration is OPEN✨
AI for Robotics workshop (4th edition): Spatial AI
🗓️Nov 21-22 Grenoble, France!
Details: tinyurl.com/bdtk2nzs
⭐⭐ 14 confirmed speakers ⭐⭐: 🧵2/3
Poster submissions (travel grant possible): 🧵 3/3
Spread the word!

VGGT for the masses 🤘! #cvpr2025

Wow, there's a "new" book from him: "Deep Learning : Foundations and Concepts". Did not know about that one --> thanks for the tip 🙏

#cvpr2025 in Paris satellite event kicking off!

Thanks so much for organizing! @davidpicard.bsky.social @vickykalogeiton.bsky.social and @matthieucord.bsky.social

BTW no IEEE ICRA account on bsky? 🙁

We talk a lot about how deep learning can enable new robot capabilities but a lot less about how we can use it to make robot capabilities more resilient and therefore make robots more deployable

Slide from an interesting talk by Prof Kostas Alexis in robots in the wild workshop at icra

Jetson @nvidia's version of our robot is available!
Compute is now on-board like a @Tesla car with FSD 🚗

Importantly, we rethink the control interface, so that you can view the video stream with the wonderful @rerundotio ⭐

Part 2 of SLAM handbook is out for public comments! let us know what you think :-) Issue tracker on GitHub awaits! Link: github.com/SLAM-Handboo...

Please share broadly to your networks so this can reach its intended audience! And we are also looking for volunteers to help instruct/mentor. Details here: cv4ecology.caltech.edu/ai-upskillin...

Introducing VGGT (CVPR'25), a feedforward Transformer that directly infers all key 3D attributes from one, a few, or hundreds of images, in seconds!

Project Page: vgg-t.github.io
Code & Weights: github.com/facebookrese...

Yaak x Hugging Face: LeRobot goes to driving school YouTube video by Yaak AI

Introducing the Learning to Drive (L2D) dataset by @yaak-ai.bsky.social & @hf.co: tinyurl.com/n4rbjehz

- 90 TB of data: 6 cameras, GPS, CAN
- Aligned w German driving task catalog w
diverse conditions 🌧️☀️❄️🚧🚦⛔️🚊
- Pwrd by NLP search: tinyurl.com/mr3hjz58
- LeRobotDataset v2.1 tinyurl.com/c4aus8f2

We're super excited to introduce the Academia Hub!

A game-changer for school, university, courses, research, projects, collaborations, publications.

Learn more: huggingface.co/docs/hub/aca...

It's a per-school offer, not individual => your faculty/admin have to apply

Go schools, go universities!

MASt3R-SLAM code release!
github.com/rmurai0610/M...

Try it out on videos or with a live camera

Work with
@ericdexheimer.bsky.social*,
@ajdavison.bsky.social (*Equal Contribution)

Deep Dive into LLMs like ChatGPT YouTube video by Andrej Karpathy

@karpathy.bsky.social has a new video on LLMs www.youtube.com/watch?v=7xTG...

I'm excited to share that we'll have Ilan Price giving a talk at the University of Cambridge on GenCast -- a state-of-the-art model for probabilistic weather forecasting 🌦️⛈️⛅️

If you're in Cambridge next Wednesday (19 Feb), consider joining us -- it's open to all!

All details (+ Zoom link) below!

Dinko Korunić - DEEP 2024 - eBPF: Features, capabilities and implementation The talk explores the powerful extended Berkeley Packet Filter (eBPF) technology. eBPF allows developers to run sandboxed programs in the Linux kernel without changing kernel source code or loading kernel modules. The talk covers key features such as network traffic monitoring, performance profiling

Dinko Korunić - DEEP 2024 - eBPF: Features, capabilities and implementation (in Croatian language) www.youtube.com/watc... #ebpf #deep2024 #security #linux #networking

The #NeurIPS2024 tutorial on “Flow Matching for Generative Modeling” is now available! Feel the learn 🤓

neurips.cc/virtual/2024...

P.S. September is a great time to visit Split (& the locations where the GoT was filmed) -- it is stil warm and not that crowded.

Home Mediterranean Machine Learning Summer School Split, Croatia 8-12 September 2025

Is there a better way to combine ML & pleasure that the Mediterranean ML Summer School 🌴?

This year it will be held in Split, Croatia 🇭🇷 on September 8-12.
Split is a beautifull city full of landmarks & opportunities to test your 3D reconstruction models in the wild..
More info www.m2lschool.org

GitHub - dkorunic/betteralign: Make your Go programs use less memory (maybe) Make your Go programs use less memory (maybe). Contribute to dkorunic/betteralign development by creating an account on GitHub.

It really made me happy that Golang betteralign (github.com/dkorunic/...) got added these days to go-recipes (github.com/nikolaydu...) as one of hot/recommended static analysis tools. Thank you! \o/
3/3

MASt3R-SLAM: easily the most robust dense monocular SLAM system I've ever seen, and also very accurate. Real-time, RGB only (no IMU), handles unknown focal length and zoom. More new videos coming soon
@ericdexheimer.bsky.social @rmurai0610.bsky.social. Paper and more: edexheim.github.io/mast3r-slam

MVImgNet2.0: A Larger-scale Dataset of Multi-view Images

Xiaoguang Han et 10 al;

tl;dr: bigger, but also better quality due to PixSfM and SAM-processing

arxiv.org/abs/2412.01430

Would you be so kind and include me? Thank you in advance 🙏