Meet Mask-bot, a robot with a 3-D human face that glows brightly enough to function in daylight. The transparent plastic face receives an internal beam of light from a projector, showing the 3-D conversion of a regular photograph. Algorithms create the voice and facial expressions.
Mask-bot can speak simple dialogue. In the video below, when researcher Takaaki Kuratate says “rainbow,” Mask-bot replies, “When the sunlight strikes raindrops in the air, they act like a prism and form a rainbow.”
Makers of the robot — the Institute for Cognitive Systems (ICS) at TU München, along with the National Institute of Advanced Industrial Science and Technology (AIST) in Japan — say Mask-bot will influence the way humans communicate with robots in the future, according to a November 7, 2011, press release.
Many robot systems use a technique that projects faces onto the front of a mask, a concept pioneered by Walt Disney in the 1960s. Kuratate said:
[Disney] made the installations in his Haunted Mansion by projecting the faces of grimacing actors onto busts.
But the inventors of Mask-bot project a human face from behind the mask. To do this, the researchers had to devise a way to beam onto the mask from the short distance of 12 centimeters. For this, they used a high-compression, x0.25 fish-eye lens with a macro adapter.
Other research groups have developed 3-D heads, but the robots appear more cartoon-like than Mask-bot, which has realistic features and expressions that are recognizable from various angles. A computer changes the face on-demand. To be classified as a robot face, Mask-bot has to function without a video image of the person speaking. The technique for animating Mask-bot is faster, the researchers say, than using many small motors to drive movement, as other inventors have done.
So how does Mask-bot get facial expressions from a photograph?
A computer system, developed by Kuratate, matches face motion data to phonemes — certain sounds. The computer then extracts facial coordinates from the selected expression. These coordinates transfer to any face image, and the computer animates the face. Emotion synthesis software helps portray a happy, sad or angry face.
To create a female or male voice, a text-to-speech system converts keyboard commands.
Mask-bot does not comprehend language; it can only listen and respond based on a fixed programming sequence. Robot researchers seek to make robots that listen and understand humans in real time and make appropriate responses.
Kuratate forsees Mask-bot as a participant in video conferences or as a companion for seniors:
Usually, participants are shown on screen. With Mask-bot, however, you can create a realistic replica of a person that actually sits and speaks with you at the conference table. You can use a generic mask for male and female, or you can provide a custom-made mask for each person.
These systems could soon be used as companions for older people who spend a lot of time on their own.
Spending time alone might be preferable to spending time with Mask-bot, but we’ll see what the future brings.
Bottom line: The Institute for Cognitive Systems (ICS) at TU München, along with the National Institute of Advanced Industrial Science and Technology (AIST) in Japan, issued a press release and video on November 7, 2011, describing Mask-bot — a robot head with 3-D imagery projected onto the inside of a mask by an internal projector and computer.