Dataset for: An integrated simulator and data set that combines grasping and vision for deep learning

To develop a simulation that collects both visual information, as well as grasp information about different objects using a multi-fingered hand. These sources of data can be used in the future to learn integrated object-action grasp representations.