paper preview
01arXiv
Stay Seated: Learning Omnidirectional Humanoid Locomotion on a Passive Mobile Chair with Casters
Teaches a Unitree G1 to move around while seated on a passive wheeled chair, using its feet to push and steer.
Paper Sweeps
Humanoid Papers from August 30 to September 5, 2026
5Locomotion1Motion Tracking1Training Infrastructure
5 papers
Locomotion
Terrain, gait, robustness, and agile movement.
paper preview
02arXiv
Learning Agile Perceptive Traversal of Sparse 3D Structures for Humanoids
Teaches a humanoid to jump up, swing across monkey bars, land safely, and duck under thin obstacles using head-mounted LiDAR.
paper preview
03arXiv
ADAPT: Agile Diffusion Action Priors for Robust and Steerable Online Text-Driven Humanoid Control
Lets people steer a humanoid with text commands while the robot keeps whole-body motions stable.
paper preview
04arXiv
MulDP: Multimodal Diffusion Policy for Autonomous Quadruped Parkour Navigation across Complex Terrains
Teaches quadrupeds to run parkour across complex terrain by combining vision, body sensing, and goal cues.
paper preview
05arXiv
World-Model-Augmented Visual Locomotion for Humanoids on Foothold-Constrained Terrain
Helps humanoids cross stepping stones, gaps, and stairs by predicting what will happen before choosing each step.
1 paper
Motion Tracking
Reference motion, imitation, retargeting, and behavior priors.
paper preview
06arXiv
Unified Motion Retargeting for Humanoids with Learned Point Cloud Correspondence
Turns human motions into humanoid-ready movement by matching body shapes and preserving the intended action.
1 paper
Training Infrastructure
Benchmarks, datasets, rewards, simulators, and tooling.
paper preview
07arXiv
Three Steps at a Time: Learning Representations from Action Sequences in Contrastive RL
Improves robot learning by judging short action sequences together, helping agents choose moves that better reach their goals.