Search papers, labs, and topics across Lattice.
AI Laboratory, University of Science and Technology of China
1
0
3
DURA reveals that visually indistinguishable adversarial patches can exploit VLA models, posing a significant threat to their deployment in real-world robotics.
