New research from the University of Waterloo has improved the way that humans communicate humanoid robots, using clever design to make interactions feel more natural.
Talking to a robot often feels stilted or delayed, because the underlying software oftens struggles to keep up with the conversation and the robot doesn’t track the person it’s speaking to.
The Waterloo researchers were able to solve how a humanoid robot can identify the direction from which human speech is coming from, re-orient itself to track that voice and speed up its conversational reaction time.
‘Creating more natural conversations with humanoid robots is an interesting challenge because speech is so key to our social interactions,’ said Ewen MacDonald, a professor in Waterloo’s Department of Systems Design Engineering.
Advertisement
The research team built the hearing system using two microphones where a human’s ears would typically sit to help generate an estimate of the direction from which audio sounds were coming. However, sounds produced by humans and other acoustic sources are reflected by surfaces such as objects or walls. A signal processing pipeline was needed to account for these reflections, which they could fool the robots into thinking the sound was coming from the wrong place.
For a robot to react as quickly as a human would sonically, its computer needs to generate an estimated location extremely quickly.The researchers developed a framework that can optimise the robot’s processing speed and characterise different sounds based on overall performance and latency. They tested the framework using recordings in a variety of acoustic environments.
According to Pranav Barot, a graduate student in systems design, the team was motivated by the desire to test the ability of humanoid robots to listen and interact with humans in real-time. One major challenge of the research was testing how the robot reorients itself to hear humans in large, loud, or crowded spaces.
‘The implications of this research are important in any scenario or environment where humanoid robots will work together with human beings, both in social robotic situations and where humans and robots are working together,’ said Barot.
The research has been published in PLOS One.