🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Catch my talk at CVPR Precognition workshop,4:15pm PT. I'll talk about some prior work on forecasting and imitative control, CVPR20 work on activity forecasting, and ICML20 work on robust planning that goes live soon stream: youtu.be/1eDWv-yS1YU site: sites.google.com/view/ieeecvf-c…

Can autonomous 🚘 identify, recover from, and adapt to distribution shifts? We play with BDL and robust control to get cars to recover&adapt when they don't know what to do At ICML with Angelos Filos Panagiotis Tigas Rowan McAllister Nick Rhinehart Sergey Levine 📄🎞️🕸️💻: oatml.cs.ox.ac.uk/blog/2020/07/0…

Our car below never saw roundabouts at training time. But using dropout ensembles' epistemic uncertainty we can choose the best worst-case plan to follow at deployment We put code online to make it as easy as MNIST to plug & play your own BDL tools: github.com/OATML/oatomobi…

Excited to announce the NeurIPS Conference workshop on 🚗 Machine Learning for Autonomous Driving! 🚗 Call for papers: ml4ad.github.io (by 14th Oct) with Nick Rhinehart, Xinshuo Weng, Daniel Omeiza, Fisher Yu, German Ros, Vladlen Koltun. #NeurIPS2020

Instead of the standard detect-then-forecast pipeline, we first forecast a video of point clouds and then run pose detection on the generated video. Forecasting point clouds requires no labels, which affords cheap scaling of the predictive component of the pipeline

RL agents explore randomly. Humans explore by trying potential good behaviors, because we have a prior on what might be useful. Can robots get such behavioral priors? That's the idea in Parrot. arxiv arxiv.org/abs/2011.10024 web sites.google.com/view/parrot-rl vid youtube.com/watch?v=AlNOow…

RL enables robots to navigate real-world environments, with diverse visually indicated goals: sites.google.com/view/ving-robo… w/ @_prieuredesion, B. Eysenbach, G. Kahn, Nick Rhinehart paper: arxiv.org/abs/2012.09812 video: youtube.com/watch?v=57h7D-… Thread below ->

Can robots navigate new open-world environments entirely with learned models? RECON does this with latent goal models. "Run 1": search a never-before-seen environment, and build a "mental map." "Run 2": use this mental map to quickly reach goals sites.google.com/view/recon-rob… 🧵>

Excited to present IC2, an agent that learns to discover and stabilize sources of uncertainty in partially-observed environments using intrinsic reward signals defined by its beliefs in a world model #NeurIPS2021 arXiv: arxiv.org/abs/2112.03899

*Deadline Extension* Call for a second round of submissions to the #ICRA2022 workshop on Fresh Perspectives on the Future of Autonomous Driving. Submit by April 25 AoE: icra2022av.org Rowan McAllister Anthony Hu Blake Wulfe Emma Saunders Felipe Codevilla Nick Rhinehart Sergey Zagoruyko

We’re thrilled to launch the 2023 Waymo Open Dataset Challenges to help accelerate research breakthroughs in fundamental AV topics, including underexplored ones such as Sim Agents. Plus, the dataset has grown with additional label types and sensor data. blog.waymo.com/2023/03/drivin…

I’m happy to share that I’ve started a new position as Assistant Professor at The University of Toronto! I'll lead the Learning, Embodied Autonomy, and Forecasting (LEAF) Lab. Find us at leaf-lab-utoronto.github.io

Interested in advancing the foundations of robot learning 🤖? Join LEAF Lab University of Toronto (Toronto, 🇨🇦). We develop ML algorithms for forecasting, planning, and enabling autonomous systems to operate safely in complex environments. Grad apps due Jan 16. Explore: leaf.utias.utoronto.ca