The field of robotics has expanded from manufacturing into healthcare, education and even entertainment. Robotics technology is helping advance humanity all around the world. While robots have typically been programmed to perform tasks by engineers, they’re now slowly being able to learn on their own.
Through technologies like artificial intelligence, machine learning and neural networks, robots are being taught to perform all kinds of tasks. As the breadth of their application grows, so does the ways they have to be taught.
Researchers at Carnegie Mellon University developed a body-tracking system to teach robots to recognize subtle body language movements. Computer vision and facial recognition technology can only go so far, so having robots understand the smaller nuances of body language helps make them better communicators. By capturing images of the human body parts in different poses and triangulating them in 3D, the researchers built a data set that robots can use to better interact with humans. Teaching robots to understand the natural human movements can make the interactions more fluid, which may have implications for industries skeptical about implementing robotics.
Written by IEEE on August 2, 2017