@geopgs Profile picture

George Papagiannis

@geopgs

PhD in Machine learning for Robotics @imperialcollege | MPhil in Advanced CS from @Cambridge_Uni

Similar User
ICARL photo

@ic_arl

Edward Johns photo

@Ed__Johns

Julen Urain photo

@robotgradient

SiQi Zhou photo

@zhousiqi9

Manon Flageat photo

@MFlageat

Luca Grillotti photo

@LucaGrillotti

Vitalis Vosylius photo

@vitalisvos19

Hannah Janmohamed photo

@hannah_janmo

Ivan Kapelyukh 🇺🇦 photo

@IvanKapelyukh

Rokas Bendikas photo

@BendikasRokas

Maytus Piriya photo

@maytusp

Seth Nabarro photo

@sethnabarro

SomeRobot photo

@SomeRobot1

George Papagiannis Reposted

Want to teach your robot new tasks from only a single demo? We've just released code for MILES, which we presented at CoRL 2024 last week. Learning is fully automated: you just provide a single demonstration, then sit back and relax! 🍹😴 Code: robot-learning.uk/miles. 🧵👇


George Papagiannis Reposted

Could robots collect and self-label their own data? It would be much easier than manually giving endless demos! New at #CoRL2024: 🦾✌️ MILES: Making Imitation Learning Easy with Self-Supervision Our robot learned to use a key from just one demo! robot-learning.uk/miles More 👇


George Papagiannis Reposted

Can a robot learn a skill with one object grasp, and then perform that skill with an entirely different object grasp? Today at #IROS2024, @geopgs will present our solution based on self-supervised learning! 🤖📷🔨 Oral at 6pm in Room 3. Come along 😀. robot-learning.uk/adapting-skills


George Papagiannis Reposted

If we train a robot to use a grasped object (e.g. a hammer), it's going to fail when the object is grasped in a novel way. In our latest paper, robots can now adapt skills to novel grasps without needing any further demonstrations! #IROS2024 See tinyurl.com/3ye6dyts. More 👇


Amazing stuff by @normandipalo

Excited to introduce Diffusion Augmented Agents (DAAGs)✨. We give an agent control of a diffusion model, so it can create its own *synthetic experience*.🪄 The result is a lifelong agent that can learn new reward detectors and policies, much more efficiently. Here's how. 👇



George Papagiannis Reposted

The future of robotics is in your hands. Literally. Excited to announce our new paper,✨R+X✨. A person records everyday activities while wearing a camera.A robot passively learns those skills. No labels, no training. Here's how. 👇


Come find us at the RSS Workshop on Lifelong Robot Learning to learn more about how we can learn robotic skills directly from a long, unlabelled human video with zero model training or fine tuning by leveraging existing Foundation Models!

🌟New paper!🌟 "R+X: Retrieval and Execution from Everyday Human Videos" By using a VLM for retrieval and in-context IL for execution, robots can now learn from unlabelled videos of humans performing tasks. No need to label and train; just *retrieve* and *execute*! More 👇



George Papagiannis Reposted

🤖 Render and Diffuse (R&D) 🤖 It unifies image and action spaces using virtual renders of the robot and uses a diffusion process to iteratively update them until they represent the desired robot actions. #RSS2024 Paper: arxiv.org/pdf/2405.18196 Project: vv19.github.io/render-and-dif…


George Papagiannis Reposted

✨ Introducing Keypoint Action Tokens. 🤖 We translate visual observations and robot actions into a "language" that off-the-shelf LLMs can ingest and output. This transforms LLMs into *in-context, low-level imitation learning machines*. 🚀 Let me explain. 👇🧵


George Papagiannis Reposted

✨Introducing 🦖🤖DINOBot. DINOBot is an imitation learning method designed explicitly around the strengths of Vision Foundation Models to fully leverage them. Thanks to that, it can learn a series of everyday skills with a single demonstration. Learn more. 👇 (ICRA 2024! 🇯🇵)


George Papagiannis Reposted

Tomorrow I will give a talk at @picampusschool covering some recent research on robotics x foundation models. Will cover topics from letting GPT-4 maneuver a robot, to building robotic foundation models like RT-X. 4PM CET. Online, register below. pischool.link/robotics

Dive into the future of #AI and #Robotics with Pi School's upcoming tech talk, Robotics x Foundation Models! Join us on Dec 14, 2023, at 4:00 PM CET for an illuminating session with @normandipalo, an AI and robotics expert. Discover the potential of merging digital intelligence…

Tweet Image 1


George Papagiannis Reposted

We’ve been showing live demos of our one-shot imitation learning method, DOME, at the #ICRA2023 exhibition. Today is the last chance to catch it, until 5pm, at stand F20. Below is a sneak preview of @GeorgePapagian1 showing how our robot can make toast! robot-learning.uk/dome


Loading...

Something went wrong.


Something went wrong.