Search papers, labs, and topics across Lattice.
6
0
6
25
Action-derived visual attention can boost robot task success rates by over 28% without relying on external labels.
A unified framework for understanding manipulation robustness could accelerate the development of robots that match human dexterity in uncertain environments.
Achieving "collect one, get one for free" in data collection, MirrorDuo drastically enhances learning efficiency by leveraging mirrored demonstrations.
State-of-the-art surgical robotics policies can be disrupted by adversarial attacks, leading to a staggering 61% drop in task success rates.
Fibration-RRT can tackle high-dimensional multi-robot motion planning by seamlessly integrating projection and decomposition strategies, proving to be both efficient and robust.
RoboLight offers the first real-world robotic manipulation dataset with linearly composable illumination, enabling systematic study of lighting's impact on robot perception and control.
