Over a million developers have joined DZone.

The AI That Can Appreciate Natural Beauty

DZone 's Guide to

The AI That Can Appreciate Natural Beauty

Now, a research team has created an AI algorithm that can understand when it is looking at something naturally beautiful in a photo.

· AI Zone ·
Free Resource

Computers have become increasingly potent at identifying and understanding images in recent years. For instance, last year, a paper was released highlighting how computers are now capable of understanding the contents of an image merely by looking at it.

It's a task that humans are incredibly good at, but without meta information, it's something that befuddles most computers — until now, at least.

Now, a team from Warwick Business School has gone a step further by training an AI system to understand when they're looking at something naturally beautiful in a photo.

Appreciating Beauty

The work, which was documented in a recently published paper, utilized deep learning to examine over 200,000 images from around the United Kingdom, with each image having been rated for its natural beauty via the Scenic or Not website.

Each image had a further array of tags that described the contents of the image. For instance, the researchers were able to use this to understand whether the picture contained a mountain or a river. They could then reference this against the scenic score for the image to understand the kind of features people typically regard as scenic.

The algorithm was then trained to start rating the images on its own.

"We tested our model in London and it not only identified parks like Hampstead Heath as beautiful but also built-up areas such as Big Ben and the Tower of London," the authors say.

Interestingly, the work uncovered that we tend to find not just obvious natural features beautiful but also certain man-made features. Buildings such as churches or bridges can add to the beauty of a scene, whereas open fields tended not to so much.

The team hopes that their work will be useful in helping us to understand the areas we live in, and especially why green doesn't always rate favorably for those living there. It could feed nicely into planning decisions when authorities aim to improve the well-being of local residents.

ai ,deep learning ,image recognition ,algorithm

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}