Robots’ maps of their environments can make existing object-recognition algorithms more accurate. John Leonard’s group
Highly accurate system learns to decipher images on its own If we’ve learned anything from
