In this paper a new dataset for robot grasping task is proposed. Compared to grasping data collected in a lab environment, the authors propose to collect the data from real world environments (homes). To collect data in the wild, the authors propose to use cheap robots (measured by the cost) with low DoF. In order to compensate the noisy behavior of the less calibrated robots, the authors model the noise as a latent variable and jointly learn it with the grasping task. Results show that the combination of the aforementioned ideas result in a robot grasping model that can work well on both lab environments, and new real world environment.

How do you make a robot toy that's both interesting and affordable? It's a hard problem: Making an interesting robot means giving it intelligence and creative autonomy, and giving a robot intelligence and creative autonomy is generally not compatible with it also being cheap. At CES a few weeks ago, we were introduced to Bots_Alive, a small company of roboticists who have managed to develop a robotic critter with a carefully thought-out animal-like personality. And by hacking an existing robot toy and using your phone as a brain, they're ready to sell it to you for 35 bucks. Bots_Alive is, essentially, selling a replacement brain for Hexbug Spider robots.