‘Eye Catcher’ Reads Audience Expressions, Triggers Face Animation
This project from the Interactive Architecture Lab is an expressive interface that utilizes emotion recognition algorithms
Eye Catcher is an expressive interface that combines industrial robotics and high power magnets with emotion recognition algorithms to read its audience and subsequently trigger the animation of a face. This installation uses technology to add emotion and context to a typically static object, surprising passersby with its movement and expressions.
The project’s principle researchers were Lin Zhang and Ran Xie from the Interactive Architecture Lab, with help from supervisors Ruairi Glynn, Dr Christopher Leung and William Bondin. The Interactive Architecture Lab is a multi-disciplinary research group and Masters Programme at the Bartlett School of Architecture at University College London.
At first glance, Eye Catcher is just an inconspicuous frame on a wall, but it comes to life through the combination of industrial robotics and high power magnets. The expressive interface uses emotion recognition algorithms to read the faces of those who pass by and trigger the animation of a face formed magnetically out of ferrofluid.
A robot running the Interactive Architecture Lab’s own Scorpion Software magnetically puppeteers the frame from behind the wall, making it move along the wall towards passersby. Those unaware of the interactive installation see unexpected movement out of the corner of their eye. When they stop and turn to look at the frame, it positions itself right in front of them.
Then, two primordial pupils rise up from the black liquid in the bottom of the frame, seemingly staring out at its viewer. A hidden pinhole camera in the frame captures the facial expressions of the onlooker and Eye Catcher responds with a range of emotions crafted out of the subtle manipulation of motion cues, creating a playful interaction as expressions are exchanged between human and interface.
You can check out Eye Catcher in action and see how it was made in the videos below: