Search papers, labs, and topics across Lattice.
4
0
3
Uncertainty quantification in VLAs can reduce the need for costly expert demonstrations by over 22%, enhancing their adaptability and reliability in real-world applications.
Removing explicit gait priors leads to a 56% reduction in energy costs and significantly improved adaptability in quadrupedal locomotion.
A hanging tray design allows robots to transport objects with minimal sliding and sloshing, revolutionizing nonprehensile transportation methods.
In-flight reinforcement learning is now feasible, with Crazyflow training a drone recovery policy in just 0.38 seconds mid-air.
