Abstract | ||
---|---|---|
We study a novel architecture and training procedure for locomotion tasks. A high-frequency, low-level network with access to proprioceptive sensors learns sensorimotor primitives by training on simple tasks. This pre-trained module is fixed and connected to a low-frequency, high-level cortical network, with access to all sensors, which drives behavior by modulating the inputs to the spinal network. Where a monolithic end-to-end architecture fails completely, learning with a pre-trained spinal module succeeds at multiple high-level tasks, and enables the effective exploration required to learn from sparse rewards. We test our proposed architecture on three simulated bodies: a 16-dimensional swimming snake, a 20-dimensional quadruped, and a 54-dimensional humanoid. Our results are illustrated in the accompanying video at this https URL |
Year | Venue | Field |
---|---|---|
2016 | arXiv: Robotics | Architecture,Simulation,Computer science,Artificial intelligence |
DocType | Volume | Citations |
Journal | abs/1610.05182 | 21 |
PageRank | References | Authors |
0.94 | 13 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Nicolas Heess | 1 | 1762 | 94.77 |
Gregory Wayne | 2 | 21 | 0.94 |
Yuval Tassa | 3 | 1097 | 52.33 |
Timothy P. Lillicrap | 4 | 4377 | 170.65 |
Martin A. Riedmiller | 5 | 239 | 23.98 |
David Silver | 6 | 8252 | 363.86 |