Communication is a lot more than just speech. It involves tone, hand gestures, eye contact, body language, and a number of similar processes. Unfortunately, because there are so many factors, it is hard to create robots that are able to truly mimic humans. But now, IBM has applied its learnings from Watson (a tech platform that unites natural language processing and machine learning to make patterns from large amounts of unstructured data) to improve robots’ understanding of human communication.
The team used video footage to help associate certain actions with phrases, and then applied this knowledge to the creation of a small humanoid robot. Robert High, of IBM Watson, demonstrated their techniques during their keynote speech at the recently concluded RoboBusiness conference.
Although the project is still in its early stages, it may dramatically increase the number of potential uses for robots. As the years pass, more and more environments will have humans and robots working together, and the improved forms of communication could be used in everything from customer service to personal assistants. However, the team has a lot more work to do before their tech can be incorporated into daily use. For example, they need to figure out how the robot’s algorithms can filter out background sounds in noisy environments.