From Text to Trajectory: How MIT’s AI Masters Language-Guided Navigation

Researchers from MIT and the MIT-IBM Watson AI Lab have developed a novel AI navigation method that converts visual data into language descriptions to aid robots in navigating complex tasks. This approach uses a large language model to generate synthetic training data and make navigation decisions based on language inputs. Although not outperforming visual-based models, [...]

from SciTechDaily https://ift.tt/CJI5kGA