Whether you prefer Van Gogh or Rembrandt, thanks to a trained neural network and 3D printing, you could soon have accurate copies of your favorite paintings hanging up in your home.
To do this, researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a system called RePaint.
This system uses a neural network to identify the layout of a painting and predict the exact color combination required to recreate it. The researchers trained the neural network to accurately predict colors using a dataset of 18,878 samples of ink stacks. The system can even adjust to the input lighting before 3D printing a copy.
Changil Kim, co-author of the paper and a postdoctoral fellow at MIT CSAIL, explains: “If you just reproduce the color of a painting as it looks in the gallery, it might look different in your home. Our system works under any lighting condition, and shows a far greater color reproduction capability than almost any other previous work.”
3D Printing Lost and Damaged Artwork
The researchers use a MultiFab multi-material inkjet 3D printer as part of the RePaint system. Capable of utilizing ten different photopolymers of differing colors, rather than the four typically available to 2D printers (CMYK – cyan, magenta, yellow and ‘key’), this provides many more shade and tonal possibilities. The final printed piece is built over several passes, with the printer layering a base UV curable photopolymer that is mixed with commercially available color pigments.
While the look of the final piece is like that of a classic oil painting, the image lacks surface texture. For example, there are no brush strokes to inspect. The researchers will continue to develop RePaint and plan to change the thickness of layers, add more colored inks to the system and produce glossy or matte finishes.
Before you get too excited about the idea of having a close to perfect Picasso in your home, RePaint can currently only create prints the size of postcards.
However, the researchers plan to use the system to recreate art which has been damaged or lost. Michael Foshey, co-author of the paper and a mechanical engineer at CSAIL, adds:
“The value of fine art has rapidly increased in recent years, so there’s an increased tendency for it to be locked up in warehouses away from the public eye. We’re building the technology to reverse this trend, and to create inexpensive and accurate reproductions that can be enjoyed by all.”
The researchers will be presenting their work SIGGRAPH Asia, in Tokyo, Japan between 4th and 7th December, 2018. Find more information about RePaint in the researcher’s paper.
Source: MIT News
License: The text of "RePaint: AI-Powered 3D Printing System Accurately Reproduces Artwork" by All3DP is licensed under a Creative Commons Attribution 4.0 International License.