The web is awash with uncredited images. The tech startup Mediachain Labs is hoping to get photographers the credit they’re due.
They’ve done this by ingesting a trove of images with Creative Commons licenses from over 30 image sharing platforms such as Flickr, along with the attribution found on those platforms. It then used neural network-powered content identification technology to de-duplicate over 400 million images, leaving it with a base of 125 million photos.
Armed with this data, users can search Mediachain’s Attribution Engine to find works simply by uploading a photo. Using the content identification technology, Mediachain can locate the work or, if it’s not in its database, suggest visually similar images that are licensed for re-use and have attribution attached to them.
Alternatively, users can search Attribution Engine for Creative Commons-licensed photos. Unlike other CC sources, however, Attribution Engine offers an embed code that enables the creators of the CC work to automatically get credit when it’s used and track where it was used (similar in spirit to how Getty has offered its free images via embed codes). In this way, Creative Commons creators can get some sense of validation and recognition for work and track its use, much as Twitter lets users track retweets and mentions and Facebook users reward posts with “likes” and shares.
However, this vision only works if a user opts to post the image with the embed code. If they simply download an image from Attribution Engine, the embed code and attribution information will be lost and the original creator won’t be able to track its use.
According to co-founder, Jesse Walden, Mediachain’s goal is to reach 1 billion images in the coming months by inviting creators to register their visual works. They’re partnering with platforms and organizations that support attribution for photographers and artists including Getty Images, The Museum of Modern Art (MoMA), Digital Public Library of America and Europeana.
At the heart of Mediachain’s approach is a technology called blockchain (popularized by cyber currencies such as Bitcoin). “You can think of a blockchain as a metaphor for a shared database, with no central point of control or authority,” Walden tells us via email. “This means the system is open, anyone can participate and build applications on top (like Attribution Engine) and users can interact and retain control of their data with no middleman.”
Unlike the blockchain databases powering Bitcoin however, Mediachain is a more decentralized database “capable of scaling at low cost, while preserving the complex relationships and revision history of structured media metadata.”
The company isn’t the first to cotton onto the idea of using blockchain to track creative works across the web. Germany’s Ascribe and U.S.-based Blackai, have released software that lets artists create blockchain ownership records for their images.
Blockchain won’t alleviate the need to file copyright notices with the U.S. Copyright Office for commercial images, but it does appear to be growing in use as a means to improve attribution in the commons.