RoboticsMachine Learning2 min reading time

HiPHI: A Large-Scale Benchmark for High-Precision Human Motion and Object Interaction

IEEE Spectrum
Read the full article
HiPHI is a new large-scale dataset capturing 617.5 hours of precise whole-body human motion, including 245.7 hours of human-object interaction, to advance humanoid robot learning. The dataset uses FrameNet to organize actions and supports training reinforcement learning policies that transfer to real robots.

More in Robotics

The Hardness Of Biology, Our Robot Friends, And The Road Ahead

Forbes
Robotics5 min read

NVIDIA Releases Nemotron 3.5 Bolt E4B 4GB: An Open Small Language Model for Local Voice Agents, Game Characters and Robots

MarkTechPost

Secretive Ukrainian Unit Behind Rapid Expansion of Deep-Strike Campaign

Bloomberg