研究Papers

A Data-Efficient Framework for Deploying Retail Humanoid Robots in the Real World

A systems-level framework for deploying VLA-based humanoids in real retail settings demonstrates that data quality matters more than data volume.

Roger Sala Sisó, Tiago Silvério, Jakob Sand +1

Zero-Collision Intersection Navigation via Compact Latent Coordination

Researchers at Ben-Gurion University developed MAPS, a hierarchical deep reinforcement learning architecture that enables autonomous vehicles to navigate uns…

Gil Lifshits, Igal Bilik, Gilad Katz

How to Train Robots That Actually Handle Novel Tasks: A Bias-Aware Data Strategy

Yu Qi, Zhang Ye, Xinyi Xu +6

A Growable Community-Driven Data Engine for Scalable Robot Manipulation

We present AXIS, a web-based, community-driven infrastructure for scalable manipulation data collection through browser-based MuJoCo-WASM teleoperation and a…

Mengfei Zhao, Dihong Huang, Yikai Tang +12

Miniature Humanoid Tele-Loco-Manipulation: VR Control + Reinforcement Learning

Researchers have created a complete system that lets an operator remotely control both the arms and legs of a miniature humanoid robot using a VR headset, ha…

Nicolas Kosanovic, Jordan Dowdy, Jean Chagas Vaz

No Training, Better Flights: Test-Time Scaling Makes Smarter UAV Navigation

Test-time scaling lets VLMs generate and refine multiple flight plans at inference, achieving SOTA navigation without any retraining.

Feinan Cheng, Dongliang Xu, Wenli Nong +4

How Haptic Calibration and Visual Feedback Improve Humanoid Robot Grasping

Low-cost calibration and visual feedback methods improve sim-to-real grasping for the NICO humanoid robot.

Juraj Gavura, Igor Farkaš

FM-VLA: Giving Robots a Sense of Touch Memory for Smarter Contact-Rich Tasks

Researchers have developed FM-VLA, a vision-language-action model that adds a dedicated force-based memory module, enabling robots to reason over sequences o…

Ruicheng Li, Qixiu Li, Ruichun Ma +8

VLMM Gives Robots a New Way to Ask "Can I Move That?" in 3D Maps

Researchers have built a mapping system that adds a motion attribute to open-vocabulary 3D scene maps, letting robots query whether objects are movable, stat…

Dibyendu Ghosh, Ayushi Shakya

Handroid: A Desktop Robot That Switches Between Hand and Humanoid

Handroid is a 27-DoF desktop robot that reconfigures from a dexterous hand into a humanoid, achieving 72% grasp success and stable locomotion.

Ruogu Li, Chenyang Ma, Sikai Li +6