With the idea of being followed, we decided to use a plain mask attached to motors (Servos) as the basis of our work.

Initially, the servos were fixed onto a platform and the mask was just attached to it on the platform itself.

We then used a face tracking code done in Processing to help us track the movement of faces within the camera’s range and we used the position of the tracked face on the screen to send instructions to the servos in terms of how much to move.

Realising that the movement is quite rigid, we decided to attach the servos together to allow more fluid and flexible movements.

 

Projection

For our projection of faces onto the mask itself, we tried just projecting a face onto the mask from a small projector.

Then we realized there will be alignment errors when the mask move. So we decided to try using the Kinect2 to aid us in the mapping of the face onto the mask well. The use of the Kinect2 helped with getting around the edges of the face, however it was difficult and unstable to translate the Kinect2’s programme to Processing, so we decided not to use it and went on to the method we used in the end.

We decided to take a uv mapping of a face and project it onto the mask. Although there are slight alignment problems initially, the image got around the edges of the mask as it was turning left and right and it has helped us achieved the feeling of the mask following you from left to right, front and down with a face on it.

 

Interface

Coming into the technical aspect, we have used the Arduino Uno board attached to 2 servo motors as our mechanism. We have also tried using the Fidget to programme the servos to work. However, as the software to code the Arduino was similar to Processing and we are more familiar with it, we decided to stick to the Arduino.

Lastly, we have tried out different ways to display our work. One was on a platform where the servos and mask are attached to the rig itself.

Then, we wanted to hang and attach it onto the wall.

But finally, we decided to use a prop that is like a container and hooked it up from the top and leave the work dangling inside. We made it more stable by attaching wires onto various points on the mask.

 

Location

Finally, at the location itself, we placed the container onto a pedestal and from a distance, the camera and projector was set up. With tweaking of the code and position, we adjusted the work to fit the distance and space of our set up and “Its alive!”

 

Issac Ting

Nathanael Goh