Project Aria @Meta

Project Aria is a research tool from Meta Reality Labs, to help build the future responsibly.

[Project Aria @Meta @meta_aria]

Privacy-first research is the core of everything we do at Project Aria. 🛡️✨

We’re thrilled to release EgoBlur Gen 2—our next-gen open-source solution for preserving privacy in egocentric video.

Key Features:

✅ Optimized for Gen 2! Tuned for high-res data & next-gen hardware ⚡️
✅ Native for VRS, MP4, JPEG, and PNG

EgoBlur Gen 2 helps you automatically detect and obscure faces and license plates on egocentric cameras from Aria Gen 2 with ease. 🔍

Dive into the models: 👉 https://t.co/DGEyDPEW9L

[Project Aria @Meta @meta_aria]

Find Meta RL-R leader Jakob Julian Engel tomorrow from 3:45pm – 4:15 pm MDT for "NymeriaPlus" at the Agents in Interactions workshop in Mile High 2A.
📷https://t.co/xIuVPiLbBv

#ProjectAria #SpatialComputing #ComputerVision #CVPR 📷

👋@jajuengel

[Homanga Bharadhwaj @ CVPR @mangahomanga]

We're thrilled to organize the 2nd Workshop on Agents in Interactions: From Humans to Robots!

Submit your best work by May 8 and join us at CVPR in Denver to discuss research in this exciting space

w/ @yufei_ye @DandanShan_ @jiaman01 @xiaolonw Alan Yuille

[Project Aria @Meta @meta_aria]

The talk is on! Join us in room 108 💡

[Project Aria @Meta @meta_aria]

Right now at #CVPR2026! 🚀

Join Meta Reality Labs Research leaders Evgeniy Olenik and David Hayden today from 1:30-2 PM MDT in Room 108 for their talk, "Aria as a Remote-Expert Platform for Developing Egocentric AI Task Assistants," at the Vita 2026 workshop. Starting now!

🔗https://t.co/TvMUJqLCC1

#ProjectAria #SpatialComputing #ComputerVision #CVPR

[Project Aria @Meta @meta_aria]

Join the Project Aria Team at EgoVis to hear the latest Project Aria Updates from Research Product Manager, James Fort.

📍Room 704/706
⏲️ 5:15-5:45pm MDT
💡Project Aria Updates!

https://t.co/Igrg97f6cH

#CVPR2026 #ProjectAria #computervision #machineperception

[Dima Damen @CVPR @dimadamen]

Join us today @CVPR rooms 704/706 #CVPR2026 at the EgoVis workshop - a long tradition of workshops on egocentric vision from #ECCV2016 10 years ago!
This edition features great keynotes, talks from relevant CVPR papers, accepted contributions&abstracts, challenge&award results.

[Dima Damen @CVPR @dimadamen]

Join us today @CVPR rooms 704/706 #CVPR2026 at the EgoVis workshop - a long tradition of workshops on egocentric vision from #ECCV2016 10 years ago!
This edition features great keynotes, talks from relevant CVPR papers, accepted contributions&abstracts, challenge&award results.

[Project Aria @Meta @meta_aria]

Meta and Project Aria are headed to Denver for CVPR 2026. ✈️

The Project Aria team will be onsite, highlighting the latest from our program on the future of Contextual AI and Robotics research! 🤖

📍 Denver, Colorado
📅 June 3–7, 2026
🤝Booth 201

We'll see you tomorrow for presentations on the cutting edge of computer vision research across CVPR's workshops and tutorials. See our snapshot agenda below for details!

[Lawrence Yunzhou Zhu ✈️ ICRA26 @LawrenceZhu22]

Excited to be at #ICRA2026 in Vienna! 🇦🇹 I'll be presenting our work EMMA during the Robot Learning I oral session (WeBT2).

🗓️ Wed, June 3
⏰ 17:35 - 17:45
📍 VIECON, Hall A3

DM me if you’d like a chat in person! 👇

[Siddhant Bansal @Sid__Bansal]

It's almost time! The EgoVis@CVPR2026 Workshop is on 3rd June!

Get ready for an amazing day of keynotes, posters, panel, and awards. Plan your day with our full schedule so you don't miss a thing!

🔗 https://t.co/vp0sOq1vMY

See you soon in Room 704/706! #EgoVis @CVPR

[Antonino Furnari @anfurnari]

Looking forward to the third edition of EgoVis workshop, next week in Denver!

If you are working in egovision, you should definitely check out our program! https://t.co/KrHmZYzIDd

We have lots of invited works from @CVPR which will make a nice reading list for your flight ;)

[Project Aria @Meta @meta_aria]

🧵Project Aria is now on Threads - just in time for @CVPR 2026! 🧵

Follow us: https://t.co/yGAKpSftCB

[Zhi (Leo) Wang @TX_Leo_Wang]

「👓Data Collection & Preprocessing with Aria Glasses @meta_aria 」

✦ Collecting — Anyone, anytime, anywhere — wear Aria glasses, do the task naturally. No calibration, no special workspace, no lighting setup. 30 min per task.

✦ Preprocessing — Aria's MPS pipeline gives us 6-DoF SLAM + 3D hand pose + synchronized egocentric RGB, all from one lightweight wearable. On top, we use off-the-shelf foundation models for lightweight, accurate object tracking — completing the hand–object pair ICT needs.

Why Aria — nothing else in the category checks both boxes today:

1️⃣ Lightweight, production-grade hardware — wearable naturally for long sessions, in any environment

2️⃣ Turnkey MPS pipeline — calibrated SLAM, hand pose, multi-stream RGB, no per-session calibration

3️⃣Accuracy isn't a luxury — our experiments show upstream pose accuracy directly bounds downstream policy performance. Aria's clean, drift-free trajectories are exactly what let HumanEgo converge on minutes of data.

🧵 6/n

[Project Aria @Meta @meta_aria]

If you can see it, a robot can do it. EgoHuman learns complex manipulation policies using only minutes of human video captured via Project Aria Gen 1 glasses.

By focusing on interaction rather than raw pixels, it transfers seamlessly across embodiments, cameras, and environments. Zero robot data required. 🚀

🔗 https://t.co/pAismnNXLA

[Zhi (Leo) Wang @TX_Leo_Wang]

1/ 🧠Humans are the best robot data source!
2/ 👓Human egocentric video is rich in quantity, but poor in quality.
3/ Beyond scaling data, smarter representation and architecture matter just as much.
4/ Want an open-source framework to train your own learn-from-human-data robot policy?

🚀We introduce HumanEgo: Zero-Shot Robot Learning
from Minutes of Human Egocentric Videos⬇️

✦ Zero-Shot Human-to-Robot Transfer
✦ Robot-Data-Free
✦ Just 30 min of data per task
✦ Collect by Anyone, Anytime, Anywhere
✦ Deploy on Any Robot, Any Camera, Any Environment
✦ Open-Source & Easy-to-Implement

Let's squeeze every bit of signal out of human data!

🌐 Website: https://t.co/JfsW8x6wtq
📄 Paper: https://t.co/tsaIiatmNi
💻 Code: https://t.co/jZjghCcjh2
📹 Video: https://t.co/QWmJmQ9GgQ
🧵 1/n

[Project Aria @Meta @meta_aria]

Another awesome partner highlight from the Project Aria research community!

⚡EgoXtreme is a novel large-scale dataset for robust 6D object pose estimation in egocentric views under extreme conditions.

Selected as a CVPR 2026 Highlight 🌟, this benchmark pushes egocentric vision to the limit. Built entirely using Project Aria glasses, the dataset leverages the hardware to capture high-fidelity first-person data in challenging environments:

⏱️ 775+ min. of egocentric video across 15 active participants.
📸 High-res 1408 × 1408 raw fisheye RGB streams (and undistorted equivalents).
🔥 Heavy motion blur, dense smoke, and 8 rapid illumination changes.
🔧Extreme scenarios including maintenance, sports, and emergency response.

Dataset: Hugging Face (taegyoun88/egoxtreme)💻 Project Page: https://t.co/3afrvNjinN

👋@taeinkwon1

[Project Aria @Meta @meta_aria]

💡 How do we build robust 3D hand tracking that generalizes across different wearable hardware?

The Augmented Vision team DFKI has taken a huge step toward solving this with EgoForce: Forearm-Guided Camera-Space 3D Hand Pose from a Monocular Egocentric Camera, just accepted at SIGGRAPH 2026.

Traditional monocular RGB methods suffer from depth-scale ambiguity and typically require laborious training on device-specific datasets. EgoForce bypasses this constraint by:

🌟 Leveraging a differentiable forearm representation to stabilize hand pose.

🫸 Utilizing a unified arm-hand transformer to predict geometry from a single egocentric view.

⚡️ Incorporating a ray-space closed-form solver for absolute 3D pose recovery.

Kudos to Alain Pagani, Christen Millerdurai, and the entire research team!👏

Website 👉https://t.co/4UbRsTaGXa
Paper 📰 https://t.co/g60BNFj82U

[Project Aria @Meta @meta_aria]

The Project Aria team is excited to be part of the Third Joint Egocentric Vision (EgoVis) Workshop at #CVPR2026! 👓✨

Don’t miss the latest Project Aria updates as we dive into the future of egocentric perception. Check out the full program of invited speakers below!

📅 June 3, 2026
📍 Colorado Convention Center | Room 704/706
🔗 https://t.co/g1FA6Cx4RU

⚡️@liuziwei7 , @mapo1 , @doughty_hazel

[Project Aria @Meta @meta_aria]

Level up your models with the Aria Gen 2 Pilot Dataset. 🚀

The Aria Gen 2 Pilot Dataset (A2PD) provides a holistic view of human-environment interaction, from cleaning and cooking to outdoor walks all on the Aria Gen 2 sensor suite.

A2PD includes:

🔹Raw sensor streams directly from Aria Gen 2
🔹Real-time machine perception outputs generated on-device
🔹Offline machine perception results produced by Machine Perception Services (MPS).
🔹Outputs from additional algorithms, including Egocentric Voxel Lifting, NVIDIA's FoundationStereo, a Speaker Diarization algorithm, and a Hand-object Interaction algorithm.

Download the data and the paper: 🔗
https://t.co/CuNZgp9vnw

#AIResearch #SpatialComputing #MetaRealityLabs #AriaGen2

[Project Aria @Meta @meta_aria]

@ericjang11 🙏 We think so too, Eric!