The Massachusetts Institute of Technology (MIT), together with the MIT-IBM Watson AI Lab, has developed a navigation method to convert visual features from images of a robot's environment into text ...