Dexterity from Smart Lenses: Multi‑Fingered Robot Skills from Everyday Human Videos

Kari Jaaskelainen

21 Nov 2025 — 1 min read

What if robots could pick up dexterous, multi-finger skills just by watching you do everyday chores? That’s the promise of AINA, a new framework that learns robot hand policies from in‑the‑wild human demonstrations captured with Aria Gen 2 smart glasses.

Why it matters: collecting robot data is hard. AINA lets anyone, anywhere record useful training data while going about normal activities—no studio rigs.

Smart glasses provide high‑res video, precise 3D head and hand poses, and a wide stereo view for depth.
AINA turns this into 3D point‑based policies for multi‑fingered robot hands.
No robot data, online corrections, simulation, or reinforcement learning needed.
Policies are robust to changing backgrounds and deploy directly on real robots.

In tests across nine everyday manipulation tasks, AINA outperformed prior human‑to‑robot approaches and includes careful ablations of design choices.

See rollouts and details: https://aina-robot.github.io • Paper: https://arxiv.org/abs/2511.16661v1

Paper: https://arxiv.org/abs/2511.16661v1

Register: https://www.AiFeta.com

#Robotics #AI #RobotLearning #ImitationLearning #DexterousManipulation #ComputerVision #Wearables #SmartGlasses #EmbodiedAI #Research

Meet GRACE: a moral governor for safer, more transparent AI

AI agents are getting powerful—so how do we make sure they do the right thing, not just the effective thing? Meet GRACE, a reason-based moral governor that keeps AI behavior aligned with human norms by separating moral reasoning from goal-driven decision-making. * Moral Module: uses deontic logic and explicit reasons

Why some fine-tuned LLMs miss phishing—and how to fix it

Not all fine-tuned LLMs spot phishing equally. A new study tests Llama 3.1 8B, Gemma 2 9B, and Mistral on high-stakes phishing detection—and uses SHAP and mechanistic interpretability to reveal why models do (or don’t) generalize. * Architecture × data diversity matters: Gemma 2 9B hits state-of-the-art performance (F1

AI that explains itself—by following the science

AI that explains itself—by following the science Black-box AI can be powerful, but it often can’t tell us why it made a decision. Concept Bottleneck Models (CBMs) try to fix that by predicting human-understandable concepts first, then the final answer. The catch: standard CBMs ignore domain-specific cause-and-effect and

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

TL;DR: Finding reliable answers in genomics is hard. GenomAgent turns one big AI into a coordinating team of specialists—and beats the current leader by 12% on a key benchmark. Genomic facts live across many databases. Standard chatbots struggle because they can’t flexibly query those sources. GeneGPT added

Read more

Meet GRACE: a moral governor for safer, more transparent AI

Why some fine-tuned LLMs miss phishing—and how to fix it

AI that explains itself—by following the science

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A