3 accepted papers
Imitation learning methods have proven effective in learning robotic tasks by leveraging multiple human-controlled demonstrations. However, existing approaches often struggle to generalize across a wide range of tasks, such as extrapolating to unseen object locations, incorporating via-point modulat
Much recent work in grasp planning has focused on data-driven approaches, using deep learning to map from images to gripper configurations. However, this approach typically fails about once per ten attempts, limiting its practicality. We sought to better understand the degree to which such failures
Artificial vision algorithms are approaching human performance in many respects. However, their computational demands are also growing, which constraints their use in mobile and time-sensitive robots. Biological vision presents a well known practical solution. Specifically, primates limit the comput