Art and machine learning

Art and machine learning

In a blog post published last June, we used visualizations to explain how neural networks work for image classification. As a byproduct, we discovered some strange and profound images produced by neural network technology.
The open source of DeepDream has aroused great interest in the machine learning and creative programming communities. Even some art trendsetters such as Amanda Peterson, Memo Akten, Samim Winiger, Kyle McDonald and Gene Kogan immediately began to explore the combination of art and machine learning.

Image: Memo Akten’s work “Google Headquarters”
After DeepDream was open-sourced, Leon Gaty, a graduate student at the University of Tübingen in Germany, published a method for using convolutional neural networks (CNNs) to decompose the style and content of images. This paper further motivated the community and artists to use neural network methods for artistic creation.

Figure: A work created by combining a photograph with a painting style (the person in the picture is Neil, a guest on an American science program)
The deep learning open source community, especially the GitXiv project, has played a great role in promoting these algorithms for art. DeepDream and style transfer algorithms were quickly migrated to a variety of computer languages ​​and software packages by the open source community, and machine learning art has been further developed.

Image: Mike Tyka’s work “Saxophone Dream”
Although machine learning and art once had little intersection, with the rapid progress in the field of machine learning, the opportunities for combining art and machine learning have increased rapidly. Both the University of London and New York University offer courses and projects that combine machine learning with art, and even the theme of the Tate Gallery IK Award in 2016 is artificial intelligence.

Everyone is welcome to join the discussion on art and machine learning.
//googleresearch.blogspot.jp/2016/02/exploring-intersection-of-art-and.html

<<:  Lighthouse effect: Why the industry is so concerned about the success or failure of the Apple Watch

>>:  Are TV games revisiting classics game remakes or just rehashing old stuff?

Recommend

Recommend 5 information flow advertising case libraries, take them for free!

It is really difficult to make information flow a...

Haobo's supercar made for China won high praise from CCTV International

The 6th CIIE showcased over 400 new products, tec...

Is your home broadband signal always insufficient? Click here!

With the popularization of the Internet, routers ...

Aiti Tribe Stories (18): The Love Between Programmers and Code

[51CTO.com original article] "Friend, can yo...

30 most common oCPC delivery questions and answers!

In this article, I have sorted out 30 of the most...

Until 3,000 years ago, the human brain was in a "split" state?

Leviathan Press: The "two-mind hypothesis&qu...

Xiaomi Clearance Festival data analysis: more youthful than rich

The Mi Fan Festival is over, and Lei Jun quickly ...