Researchers from UC Berkeley and Siemens designed a robot, trained with our GPUs, that learns how to grip new objects just by studying a database of 3D shapes.

Robot Uses Deep Learning to Grasp Awkward and Unusual Objects

Using a GTX 1080 GPU and cuDNN with the TensorFlow deep learning framework, the team generated 6.7 million synthetic point clouds from thousands of…


Source

Leave a Reply