Add to Favourites
To login click here
AI researchers have developed a new vision-based algorithm that enables robotic dogs to autonomously scale high objects, leap across gaps, crawl under thresholds, and squeeze through crevices. This algorithm, which is based on a simplified machine learning technique, was tested on two off-the-shelf robots and the results were impressive. The algorithm combines perception and control, using images from a depth camera and machine learning to process inputs and move the robot’s legs.