1/8 Meta's SAM2 has some really fun capabilities. Coolest one I found ? Tracking through a pretty wacky transition. All this with just a single point prompt on frame 0. Frame-by-frame breakdown and more examples in the 🧵 below!

7/8 Finally, had to try Transformers (wink-wink)! Several point prompts but all before the transition through the underpass! Roughly - Point prompts for Bumblebee on frame 050, on the Decepticon on frame 089 (before it disintegrates!), a couple on the bridge on frame 175.

8/8 Credits - Used Meta's (surprisingly snappy) web demo. Video sources below. [YT] - youtu.be/_57Opqw6Kkc?t=… - youtube.com/watch?v=udKE1k… - youtu.be/_Ufcn3YC-Js?si… [Pexels] - pexels.com/video/person-v… - pexels.com/video/travelin… - pexels.com/video/view-of-…

This is very cool! TLDR; DUSt3R predicting pointmaps in global coordinate frame (eliminating need for BA). A very neat way and implicitly SLAMy. hengyiwang.github.io/projects/spann…

RRC paper "Open-Set 3D Semantic Instance Maps for Vision Language Navigation (O3D-SIM)" has been accepted in Advanced Robotics! journal Work led by Laksh Nanwani, Kumaraditya Gupta, Aditya Mathur, & Swayam Agrawal, under guidance of Prof. Madhava Krishna & Dr. A. H. Abdul Hafez.

Places are composed of things. Recognizing & retrieving these things instead of the whole image enables 🧭viewpoint invariance 🖼️semantic interpretability 🔮open-set recognition 🧵on our #ECCV2024 paper: Revisit Anything: Visual Place Recognition via Image Segment Retrieval 👇

Here’s our new #VPR take on the good old ‘revisiting problem’ in #SLAM / #localization #ECCV2024 European Conference on Computer Vision #ECCV2026 Incredible effort by Shubodh Sai and Kartik Garg Catch Shubodh Sai in person at the venue ILR Workshop: 30 Sep, 10-11am CEST Main Conf: 3 Oct, 10:30am-12:30pm CEST

Quick #success story of #PhD #anxiety test: ✅get paper accepted #ECCV2024 ✅get visa approval on Day 2 of European Conference on Computer Vision #ECCV2026 ✅✈️ UK to Italy on Day 4 (presentation) ✅collect poster minutes before the presentation slot ✅ready to answer tough Qs Catch ahmad khaliq Oct 2, 16:30

Happening now European Conference on Computer Vision #ECCV2026 at Poster Board ID: 244 if you are attending #ECCV2024 Shubodh Sai is presenting #RevisitAnything - he is also looking for #PhD / #research opportunities

It's time for multilingual Indic Vision Language Models ! Happy to present evikrAI, an Indic-flavored technology to assist e-commerce applications. x.com/BharatGen_tech…

🚀 Introducing Frank 🤖—a whole-body robot control system for day-to-day household chores. Frank has been in works for the past year—a tightly coupled hardware and a remote teleoperation interface co-led by Michelle, Nandini, and Pulkit Agrawal. Thanks to Aditya Agarwal and

Had a lot of fun recording these videos! 🥳 Hands down the best teleoperation system that I have tried out till date! 🫡

How can robots compositionally generalize over multi-object multi-robot tasks for long-horizon planning? At #CoRL2024, we introduce Generative Factor Chaining (GFC), a diffusion-based approach that composes spatial-temporal factors into long-horizon skill plans. (1/7)

We figured out a way to solve long-horizon planning problem by composing a bunch of modular diffusion models in a factor graph! This allows us to reuse the diffusion models in unseen new tasks and achieve zero-shot generalization to multi-robot collaborative manipulation tasks.

Presenting EgoMimic at #CoRL2024! 🎉 Effortless data collection with Project Aria @Meta glasses—just wear & go. Our low-cost manipulator leverages this scalable data to perform grocery handling, laundry, coffee-making & more. Thrilled to be a part of this effort! egomimic.github.io

We started this moonshot project a year ago. Now we are excited to share our progress on robot learning from egocentric human data 🕶️🤲 Key idea: Egocentric human data is robot data in disguise. By bridging the kinematic, visual, and distributional gap, we can directly leverage

Introducing DART: breaking the barriers for robotic data collection by enabling anyone, anywhere in the world to control robots without even having a robot. Just log into dexhub.ai to contribute and control robots at much lower fatigue and higher speed than

🚀Excited to share SceneComplete: an open-world 3D scene completion system for constructing a complete, segmented 3D model of a scene from a single RGB-D image.🖼️🤖 SceneComplete enables dexterous grasping and robust robot manipulation in highly cluttered scenes - a short 🧵

At the recently concluded International Conference on Intelligent Robots and Systems (IROS 2024) that was held in Abu Dhabi, IIITH’s Robotics Research Centre made a splash with 6 research papers. Here’s a brief roundup of some of that cutting-edge work. blogs.iiit.ac.in/iros-2024/

Thanks for featuring our work Reality Labs at Meta Project Aria @Meta AI at Meta. We show that embodied human data is a scalable data source for robot learning. Check out the full Meta AI blog post covering our story -ai.meta.com/blog/egomimic-…