Personalized Aesthetics: Recording the Visual Mind using Machine Learning

Originally published at:

Visual aesthetics are very personal, often subconscious, and hard to express. In a world with an overload of photographic content, a lot of time and effort is spent manually curating photographs, and it’s often hard to separate the good images from the visual noise. The question we put forward at EyeEm is: can a machine…

I'm having fun wading through the pedantic pseudo-analysis here. The topic is similar to my undergrad thesis, which suggested that we're a long way off from using standard CNN models to partition images based on single user's preferences in a art space detached from reality like abstract art. Maybe we just need more training data.

Great work. Finally this is picking up. Worked on Computational Aesthetics about 12 years ago but limited it to Early Vision and dropping semantics.

Great work! We are trying to solve a similar problem with Cornea Ai ( albeit more from the point of view of virality and popularity than aesthetics. Trick is to keep adding to the dataset more and more variety of photos accounting for demographics, age and gender apart from events and places.

nice post i already bookmarked it