Meet MapAnything – a transformer that directly regresses factored metric 3D scene geometry (from images, calibration, poses, or depth) in an end-to-end way. No pipelines, no extra stages. Just 3D geometry & cameras, straight from any type of input, delivering new state-of-the-art

Simulation drives robotics progress, but how do we close the reality gap? Introducing GaussGym: an open-source framework for learning locomotion from pixels with ultra-fast parallelized photorealistic rendering across >4,000 iPhone, GrandTour, ARKit, and Veo scenes! Thread 🧵

Tune in @ ICCV on Mon @ 10.30am where I talk about everything 3D + Realism: - Hyperscape: Gaussian Splatting in VR - FlowR: Flowing from Sparse-2-Dense 3D Recon - BulletGen: Improving 4D Recon with Bullet-Time Gen - MapAnything: Universal Feed-Forward Metric 3D Recon 🧵👇

Excited to share our latest work from the ByteDance Seed Depth Anything team — Trace Anything: Representing Any Video in 4D via Trajectory Fields 💻 Project Page: https://t.co/Q390WcWwG4 📄 Paper: https://t.co/NfxT260QWy 📦 Code: https://t.co/r2VbOHyRwL 🤖 Model:

Games are all you need? 👀

Check out the blog from @CarnegieMellon @CMU_Robotics @roboVisionCMU covering MapAnything! Stay tuned on our GitHub repo for some exciting updates soon 🤩

Researchers from the Robotics Institute + Meta Reality Labs have built a model that reconstructs images, camera data or depth scans into 3D maps within a unified system! MapAnything captures both small details and large spaces with high precision 🌎🗺️📌 https://t.co/0dSLjzfeYc

FlowR: Flowing from Sparse to Dense 3D Reconstructions @TobiasRobotics et 8 al with @Nik__V__ tl;dr: Mast3r -> 3D rec -> render new view -> flow matching & refine reconstruction with synth views -> better rec https://t.co/fmTkC15rSl

three years ago, DiT replaced the legacy unet with a transformer-based denoising backbone. we knew the bulky VAEs would be the next to go -- we just waited until we could do it right. today, we introduce Representation Autoencoders (RAE). >> Retire VAEs. Use RAEs. 👇(1/n)

Meta Reality Labs is looking for research interns! ✨ If you're excited about building world models for AR/VR and immersive experiences, come join us. 🙌 Roles available across Europe (I’m based in Zurich 🇨🇭) and the US. You can DM me or find me at #ICCV to chat. 🙂 @AIatMeta

@jianyuan_wang I guess you tried the DAv3 style init startegy for VGGT back in December'24 with the main difference being that it was full alternating (ablation c) instead of the partial one (last layers only) DAv3 uses?

Interesting ICLR submissions 🤩 Depth Anything 3 - My TLDR: Init multi view transformer of VGGT with later layer DINO weights and use teacher model trained on synthetic data only for pseudo labelling real world datasets https://t.co/1gq6WSZM5c Trace Anything - My TLDR: VGGT

Looks like researchers in visual localization are good reviewers: I see so many familiar names in the list of ICCV outstanding reviewers! Congrats to @alessiodelbue @masone_carlo @kgcs96 @jcivera @Parskatt Julian Kooij @maththrills @NicStrisc @Nik__V__ (and me, happy to be in

Glad to be recognized again as an outstanding reviewer! 🤗

We introduce RAVEN, a 3D open-set memory-based behavior tree framework for aerial outdoor semantic navigation. RAVEN not only navigates reliably toward detected targets, but also performs long-range semantic reasoning and LVLM-guided informed search

📢 SceneComp @ ICCV 2025 🏝️ 🌎 Generative Scene Completion for Immersive Worlds 🛠️ Reconstruct what you know AND 🪄 Generate what you don’t! 🙌 Meet our speakers @angelaqdai, @holynski_, @jampani_varun, @ZGojcic @taiyasaki, Peter Kontschieder https://t.co/LvONYIK3dz #ICCV2025

Show-off your own Hyperscapes! Thead 🧵 below 👇 showing off various user-created Hyperscapes! They all look SOOOO good! Add your video in the comments or @ me, and I will add it to the thread!

🎉CameraBench has been accepted as a Spotlight (3%) @ NeurIPS 2025. Huge congrats to all collaborators at CMU, MIT-IBM, UMass, Harvard, and Adobe. CameraBench is a large-scale effort that pushes video-language models to reason about the language of camera motion just like

Introducing: Hyperscape Capture 📷 Last year we showed the world's highest quality Gaussian Splatting, and the first time GS was viewable in VR. Now, capture your own Hyperscapes, directly from your Quest headset in only 5 minutes of walking around. https://t.co/wlHmtRiANy

Awesome thread; always fun to see @ducha_aiki stress testing things!

Forget pointmaps, pinhole, image-only, up-to-scale… and meet MapAnything! Truly amazing, long-term effort led by @Nik__V__ that we‘re super excited to share today!