Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Reusable neural skill embeddings for vision-guided whole body movement and object manipulation

Nov 15, 2019
Josh Merel, Saran Tunyasuvunakool, Arun Ahuja, Yuval Tassa, Leonard Hasenclever, Vu Pham, Tom Erez, Greg Wayne, Nicolas Heess

Share this with someone who'll enjoy it:

Both in simulation settings and robotics, there is an ambition to produce flexible control systems that can enable complex bodies to perform dynamic locomotion and natural object manipulation. In previous work, we developed a framework to train locomotor skills and reuse these skills for whole-body visuomotor tasks. Here, we extend this line of work to tasks involving whole body movement as well as visually guided manipulation of objects. This setting poses novel challenges in terms of task specification, exploration, and generalization. We develop an integrated approach consisting of a flexible motor primitive module, demonstrations, an instructed training regime as well as curricula in the form of task variations. We demonstrate the utility of our approach for solving challenging whole body tasks that require joint locomotion and manipulation, and characterize its behavioral robustness. We also provide a high-level overview video, see .

   Access Paper Source

Share this with someone who'll enjoy it: