Combining Sound and Vision into Robotic Perception

Meet Popeye - the robot who hates spinach.

Robotic and electronic perception has changed quite a bit over the past few years.  Digital cameras and camcorders can recognize faces and smiles, CCTV systems try to find crime before it happens, voice recognition software is constantly getting better and toy pets can offer simple interactions with its users.  All of this technology is relatively primitive when compared to human perception.  Robots are able to respond to simple commands or gestures, but are unable to comprehend the subtle nuances of spoken and visual communication or follow a command when they are in a busy environment filled with a multitude of sounds and visuals.

This is all starting to change thanks to a European project named Perception on Purpose and their robot, Popeye.  The goal of the project was to develop an artificial intelligence system that is able to locate who is giving a command by combining audio and visual clues.  Project coordinator Radu Haroud explains:

"It is not that easy to decide what is foreground and what is background using sound alone, but by combining the two modalities – sound and vision – it becomes much easier," Horaud continues. "If you are able to locate ten sound sources in ten different directions, but if in one of these directions you see a face, then you can much more easily concentrate on that sound and throw out the other ones."

This project, as well as the Mind Machine Project, are looking to make waves in AI.  Perception on Purpose will give a robot the ability to locate a single target - probably the person yelling expletives at it (that would be me).  The science behind this project is the idea that someone being able to understand what to focus on in a busy room and that person's cognitive intelligence are associated with each other.  Therefore, an AI system should be able to make an educated guess about its environment based on previous intelligent decisions and experiences.

Artificial intelligence is starting to evolve into something truly frightening.  As with most AI projects, Perception on Purpose is being researched in an attempt to develop robots that will be able to help those who cannot help themselves.  AI is slowly gaining more human abilities that are based on more than algorithms and assumptions.  Robots will soon have the ability to learn, comprehend and adapt.  Add in the future of extreme speed processing based on quantum computing, and I fear that the robot armageddon is being accelerated.

The final results of the project will be given at CogSys 2010, January 27 & 28 in Zurich.  You can download a PDF of the final report here.

::from Wired UK::
::thanks to My Outsourced Brain for the link::