30 Jun 2015 Picsbuffet – Images Arranged by Similarity on a Zoomable 2D Map
Picsbuffet is the latest work of Dr. Kai Barthel of HTW Berlin and Chief Visionary Officer (get it?) of Pixolution, a company providing technology and research around visual search solutions in the stock photo industry, and others.
It is a “Google Maps style” image search platform for stock photos, utilising Pixolution’s visual search technologies to show a 2D ‘map’ of images arranged by relevance – a fused measure of keywords and visual similarity.
It contains one million images sourced from Fotolia – still the most open and functional API in the market. Kai says they “could and will” add more images to their database, once they’ve locked down a server solution powerful enough to efficiently sort large volumes of files.
The relevancy algorithm does a great job of clustering related topics side by side, but you can at times find some ‘hard edges’ of unrelated images. Asked about this, Kai explained this is due to the quantity of relationships between different topics. Preserving all image relationships is impossible on a 2D map, but he’s working on changing the format to a ‘graph’ of images to address this issue.
Picsbuffet is an academic experiment, though with obvious commercial implications for any business that has visual content, including product images from manufacturers such as jewellers and fashion labels. And, of course, stock photo agencies.
Check it out and let me know what you think in the comments.