Robot peels banana using deep learning
Researchers from Japan have developed a machine-learning system that allows a robot with two arms to pick up and peel bananas.
Researchers from the University of Tokyo developed a machine-learning system that allows a robot with two arms to pick up and peel bananas.
Robots have difficulty manipulating deformable objects like soft fruit because they are unable to model them and lack knowledge of the force and dexterity required.
Heecheol Kim is a researcher at the Intelligent Systems and Informatics Laboratory of the University of Tokyo. He developed a system using \”goal-conditioned deep imitation learning\” (DIL). This system can be taught dexterous manipulative skills by using data from human demonstrations.