Sunday, September 6, 2015

In the style of the Master

Computers can be made to do complex tasks, as long as we are able to break up the complex task into simple, direct, logical instructions. For machines that solve problems in medicine, astrophysics or geology, it was just a matter of time before somebody broke up the process of fine art into the constituent activities of composition, colouring, blending, accentuating and so on, and got the machine to create art.


The materials employed, the brush strokes applied, the process of building layer upon layer of paints, all are unique to a given artist. For the Masters of the art, these processes have been studied in detail. Also, there is a large enough body of work by each such artist to lend itself to research, and to draw conclusions about strokes and resultant patterns.


The inevitable has happened. A computer, after an input of a suitable algorithm, took about an hour to create this self portrait of Van Gogh.

(http://www.hitc.com/en-gb/2015/09/03/computer-algorithm-recreates-van-gogh-painting-in-one-hour/)


Computers are not composing original pieces of art work yet. The research by Gatys, Ecker and Bethge (2015) into instructional algorithms for a computer, distinguishes between the content and composition of a given art work, and the style of execution. The computer was able to finish a given composition in the style of the artist specified.
(http://arxiv.org/pdf/1508.06576v1.pdf)




















All artists have their favourite themes, and Van Gogh painted several compositions with cypresses.
They were painted at different times of the day, with different hues in the skies, and with different "focal points". There are two women in the first painting shown here, and two men in the second painting.

While the authors, Gatys et al, have chosen a few well loved and recognised paintings for their research, and development of their algorithm, it will be interesting to see how the computer deals with such paintings on the same theme and by the same artist. 

Friday, May 22, 2015

Kinetic Art

A number of interesting sculptures and installations are made with "found objects" and I got to see a number of them at the Bay Area Maker Faire, at San Mateo, on 16th and 17th May. The focus in a Maker Faire is on do-it-yourself, on innovation, and on technology. This one was no exception, and had a whole lot of fresh ideas, products, and people.

I specially went to visit the Kinetic Art section. There was a lot of use of electronics to create movement, lights, sound or animation. And often the creations were quite dramatic!



This composition shown above looked metallic, yet fragile, and certainly seemed mobile, and with a mind of its own. It was like a giant insect! The coloured one with two coloured blades, seen behind was another piece, and looked like it may take off any moment!   


                                                    
This sculpture, again composed of found pieces, looked like an animated dancer, and you could imagine it dancing to some rhythmic music. The artist obviously has a huge collection of these spare pieces of wood, metal, glass and plastic, and uses them as per his imagination. All the pieces shared here were by Nemo Gould, an American artist, known for his kinetic "found-object" sculptures.


There were several pieces of Gould's, which used a fair amount of electronics as well. For example, there was a complex electronic "cuckoo clock" shown here.

In making these objects, artists have to brush up or learn on several skills like those of welding, glass blowing, soldering, sawing and on various materials.




That is where the appeal of events like the Maker Faire lie: in exposing a generation used to pushing and clicking buttons, to a whole lot of mechanical work with their hands.

Processes on cloth like stitching, knitting, quilting, crochet and embroidery all found their place at the Faire, and seemed to attract many young new learners. The events will no doubt keep growing each year!