Motion Blur

Just found some fun images while cleaning up some old files. The motion blur makes the gestures of the performer hang in the air, communicating much more information than a simple static image.

Motion-blurred images of the T-Stick in performance
Alexander Refsum Jensenius uses a more sophisticated video-processing technique to produce motiongrams and videograms, which he uses for analysis of musical performance. You can read about them in his PhD Dissertation.

Is research in your future?

Look at the bottom of the poster on the left…

The “Is research in your future” tag line is attached to the Gyrotyre, a DMI created by Elliot Sinyor in the IDMIL. Marcelo Wanderley (my PhD advisor) is holding it, and Stephen McAdams and I are holding T-Sticks. The pic is from last year’s event; another one made it to the McGill home page:

Fun photo manipulation

I spent a few minutes the other day messing around with the online face transformer from the University of St Andrew’s Perception Lab. It allows you to upload a photo and “transform” it into someone older, younger, an ape, etc… Personally, I enjoy the results best when the system is cruelly abused, and given faulty eye and mouth positions – here’s a before and after:

Mango sensors


During my endless search for digital compasses for use in a current project, I recently found a reference to an entirely original technique for sensing direction:

The discrete sensing elements employed in such sensor modules are generally called magnetometers and/or mango-inductive sensors.

(emphasis mine) I guess it senses the Earth’s mangetic field ­čśë

Seriously though, good/cheap/fast digital compasses are hard to find – I’ve been playing around with the Robot Electronics CMPS03, but it’s crazy power-hungry and not tilt-compensated. I’ve got some Honeywell HMC6343’s with breakout boards on the way, but the datasheet says they max out at 10Hz. I’ll probably just use accelerometers for fast movements, and calibrate continuously to the slower compass.