amwatson

15 Mar 2015

This isn’t really “art” per-se, but a piece of research I really like, that I think has has cool artistic implications.   Building Rome in a Day is a research project that can take a sufficient (though not as many as you might think) Flickr photos of a particular site and, with minimal geo-data, create a decent 3D reconstruction in less than 24 hours.  This means that with a collection of photos from a particular site, one could create a plausible 3d image without having to take their own photos.

I think you could do a lot of interesting things with this research.  It would be interest to construct one of these models, and then explore it with a virtual reality headset.  With the advent of WebVR, it might be possible to create an interface where 3D models can be hosted and perused.  I think the ability to not only do this, but to do it quickly and with this level of success should be interesting to anyone interested in doing 3d work with real-world models.

It’s worth pointing out that, while a lot of work is done to ensure the compute power used isn’t exorbitant, the researchers still use far more than what one might consider “commodity” hardware.  It’s not reasonable to expect a consumer to make use of this algorithm without paying a heavy fee, or stretching out the reconstruction time.  I wonder what sort of implications this algorithm might, have, though, and if something cool might be doable on a smaller scale?

Building Rome on a Cloudless Day is an offshoot of the Building Rome in a Day project, which aims to have the same results/performance without a distributed system.

 

La Maison Sensible is an interactive installation in which the walls and furniture are given sensation and emotion.  Projections on the space respond to the user’s movement and touch, giving the sense that the “house” is responding to their actions with empathy and emotion.  I’m interested in this installation, because it uses technology and interaction to make the room itself into a character, which would be difficult to pull of without computation.  If I had a hand in this piece, I would want to give a narrative to the installation — have the room’s relationship with the audience change, or reveal something about its history.  Perhaps a user’s actions reveal some “memory” of the room and cause it to become sad or angry.  I think a truly “living” interactive room has a lot of potential as a setting and start of a performance.