Grasping Student: semi-supervised learning for robotic manipulation
This addresses the data efficiency problem for robotic manipulation researchers and practitioners, offering an incremental improvement over existing methods.
The paper tackles the bottleneck of gathering real-world robot data for grasping by developing a semi-supervised system that uses unlabeled product images to enhance learning from a small sample of robot experience, achieving performance equivalent to a 10-fold larger dataset in low-data regimes.
Gathering real-world data from the robot quickly becomes a bottleneck when constructing a robot learning system for grasping. In this work, we design a semi-supervised grasping system that, on top of a small sample of robot experience, takes advantage of images of products to be picked, which are collected without any interactions with the robot. We validate our findings both in the simulation and in the real world. In the regime of a small number of robot training samples, taking advantage of the unlabeled data allows us to achieve performance at the level of 10-fold bigger dataset size used by the baseline. The code and datasets used in the paper will be released at https://github.com/nomagiclab/grasping-student.