These results validate the performance of aerial grasping based on our proposed wholebody grasp planning and motion control method. However, for most vehicles, high performance over rough terrain reduces the travel speed and/or requires complex mechanisms. We extend GPS in the following ways: (1) we propose the use of a model-free local optimizer based on path integral stochastic optimal control (PI2), which enables us to learn local policies for tasks with highly discontinuous contact dynamics; and (2) we enable GPS to train on a new set of task instances in every iteration by using on-policy sampling: this increases the diversity of the instances that the policy is trained on, and is crucial for achieving good generalization. To increase the spike decision rates, iterative spiking training with actual blockers is required.
We're told that Markus bet that this thing could only work in theory, and lost: This video introduces the monospinner, the mechanically simplest controllable flying machine in existence. With the company's ultra-low power, high performance Myriad 2 processor inside, the Fathom Neural Compute Stick can run fully-trained neural networks at under 1 Watt of power. As a tinkerer and builder of various robots and flying contraptions, I've been dreaming of getting my hands on something like the Fathom Neural Compute Stick for a long time. Last year in Seoul, KAIST's Unmanned Systems Research Group participated in an autonomous car demo in downtown Seoul.