FLORE learns force-reactive humanoid loco-manipulation from handheld demos
On three contact-rich robot tasks the force-reactive policy succeeds in 80% of rollouts versus 26.7% for position-only.
Stories per day · 30 days
3 RSS
Relatedrobotics2computer-vision1diffusion-policy1force-control1loco-manipulation1pose-estimation1self-training1
19 stories, newest first · one card per story
Stories tagged #humanoid
HumanoidPose estimates robot joint pose from video alone
On three real humanoid platforms, joint-angle error fell 18-32% versus the best baseline on each.
λ_0 humanoid model uses 500 hours of human loco-manipulation data
Zero-shot transfer to Unitree G1 reached 90.0% on SIMPLE and 62.5% on real loco-manipulation tasks.
FAITH: Feasibility-Aware Safety-Filtered RL for High-Dimensional Systems
FAITH filters robot actions with a learned safety value, preserving high task return without feasible-start safety violations.
VioLA: Learning Generalist Humanoid Control Policies from Human Data
VioLA trains a generalist humanoid policy mostly on human motion and follows real-robot instructions without task fine-tuning.
AI breakthroughs in robotics won’t change your life any time soon
Experts say Tesla Optimus and humanoid hype overstate how soon AI will handle real-world physical work.
Reflex: Visually Grounded Reactive Humanoid Control
Reflex enables a Unitree G1 to catch thrown boxes from egocentric RGB-D, with 65% success at 2.5 meters.
QF3: Fast Flow RL with Filtered Q-Gradients
QF3 trains robot flow policies off-policy, learning humanoid locomotion from scratch about ten times faster than FPO++.
Can Safeworld convince people that gen AI robots won’t hurt them?
Safeworld exits stealth with over $12 million to simulate and underwrite generative-AI robot safety.
InterMimicGen: Scaling Humanoid Loco-Manipulation through Self-Evolving Motion Imitation
InterMimicGen grows humanoid loco-manipulation data by retargeting human demos and self-evolving a physics tracker.
CoDance: Learning Reactive and Compliant Human-Humanoid Interaction from Video
CoDance learns compliant two-hand humanoid dancing from one partnered-dance video, including on a physical robot.
InterEvolve: Test-Time Evolution of Reward Programs for Humanoid Loco-Manipulation
InterEvolve evolves reward programs at test time so humanoid controllers solve new loco-manipulation tasks without retraining.
CrossBFM: Distilling a Shared Latent Behavior Space Across Humanoid Embodiments
CrossBFM distills one latent behavior space that transfers across humanoid robots in under an hour of GPU training.
Counterfactual Video Generation Enables Scalable Humanoid Loco-Manipulation
PRISM turns a few real videos into counterfactual data so a humanoid carries novel objects without real-world fine-tuning.
OTRetarget: Joint Robot and Object Motion Retargeting via Optimal Transport
OTRetarget jointly retargets humanoid and object motion from human demos using entropic optimal transport.
Systematically Exploring the Capabilities of GPT-6 Astra as Embodied Policies
A study finds GPT-6 Astra useful for robot task decisions but unreliable for direct physical control.
REALM: A Coarse-to-Fine Generative Framework for Embodied Reactive Listening
REALM generates reactive listener facial motion with coarse trajectories plus stochastic expression residuals, including Ameca deployment.
ForgetMimic: Motion Unlearning for Reinforcement Learning Humanoid Control
ForgetMimic unlearns selected humanoid motions on Unitree robots while keeping other behaviors.
Toyota claims plan for 400,000 factory robots won’t replace human workers
Toyota plans to invest $6.42 billion a year from 2028 to deploy 400,000 factory robots.