The most frustrating part of imitation learning is collecting huge amounts of teleop data. But why teleop robots when robots can learn by watching us?. Introducing Point Policy, a novel framework that enables robots to learn from human videos without any teleop, sim2real, or RL.

We are extending the paper submission deadline to August 15, 2025 23:59 AOE. Submit your latest works on generalizable robot priors to the #CoRL2025 #GenPriors workshop!.Webpage:

Call for Papers: Workshop on.Generalizable Priors for Robot Manipulation at #CoRL2025 🤖. 📅 Deadline: Aug 5. Submit your recent works on generalizable robot manipulation and come participate in our workshop!.

RT @ShivaamVats: Is "scaling is all you need" the right path for robotics?. Announcing our @corl_conf workshop on "Resource-Rational Robot….

RT @DJiafei: 1/ 🚀 Announcing #GenPriors — the CoRL 2025 workshop on Generalizable Priors for Robot Manipulation!. 📍 Seoul, Korea 📅 Sat 27 S….

Lastly, kudos to the organizing committee for the workshop!.@mlevy1221 @DJiafei @IvanKapelyukh @abhishekunique7.

Call for Papers.• 4‑8 pp (CoRL format, non‑anonymized).• Submit via OpenReview (.Whether it’s new algorithms, datasets, or insights on scaling generalist robot policies—submit your latest work on scalable priors! .Paper deadline: Aug 05, 2025.

Invited speakers span academia & industry:.@leto__jean , @AjayMandlekar , @xiaolonw , @RanjayKrishna , @gao_young , @GeorgiaChal ,.@haroldsoh , @Ed__Johns . Expect insightful talks & a lively panel!.

Relevant topics:.• Bottlenecks to OOD generalization.• Leveraging vision & language pretrained models.• Sim‑to‑real with generative pipelines.• Object‑centric & 3D representations.• Benchmarking frameworks for generalist skills.Got fresh ideas? We want to hear them!.

Why priors? Generalist robot policies must master diverse tasks, domains & data sources. We’ll ask the big question: what priors let policies generalize with precision & reliability?.

Announcing #GenPriors — the #CoRL2025 workshop on Generalizable Priors for Robot Manipulation!.📍 Seoul, Korea  📅Sat 27  Sep  2025. The workshop focuses on building generalist robot policies that work across a wide variety of environments. 🔗

RT @IlirAliu_: Robots no longer have to choose between being precise or adaptable. This new method gives them both!. [📍 Bookmark Paper & Co….

RT @LerrelPinto: It is difficult to get robots to be both precise and general. We just released a new technique for precise manipulation t….

RT @Zifan_Zhao_2718: 🚀 With minimal data and a straightforward training setup, our VisualTactile Local Policy (ViTaL) fuses egocentric visi….

RT @chris_j_paxton: A nice pipeline: use a VLM to find objects in scene, get close, and use a well-constrained visuo-tactile policy to hand….

RT @Raunaqmb: Generalization needs data. But data collection is hard for precise tasks like plugging USBs, swiping cards, inserting plugs,….

Robots can now perform contact-rich tasks in your kitchen, office, or factory—without retraining. This project was led by the amazing @Zifan_Zhao_2718, and a fun collaboration with @cui_jinda_hri, @Raunaqmb, and @LerrelPinto!.

We have open-sourced the code. Try out ViTaL on your robots!. Project page: Arxiv: Code:

Since the learned ViTaL policy only uses egocentric vision and tactile sensing, we can also deploy these policies on new robots in new environments.

ViTaL achieves >90% success on a range of precise, contact-rich tasks like plugging USBs, swiping credit cards, inserting plugs, and key insertion -- in cluttered, unseen environments.

Key insights:.1⃣Tactile sensing is essential: Removing it drops success by 40%—vision alone isn’t enough for fine manipulation. 2⃣Semantic visual augmentation: Trains policies to ignore background clutter. 3⃣Residual RL: Boosts precision without sacrificing generalization.