Last week CEPIC, the “centre of the picture industry”, held its annual Congress in Juan-les-Pins in France. This was the second time that the event was held at the Palais du Congres in Juan-les-Pins near Antibes, which is proving to be a great venue for the Congress with many repeat visitors who also attended last year.

IPTC was well represented at the event, with Managing Director Brendan Quinn presenting on two panels and many other IPTC members involved in presentations, panels or simply attending the event. IPTC members either presenting or in attendance included Google, Shutterstock, Getty Images, ANSA, IMATAG, PA Media / Alamy, TT, dpa, Adobe, Xinhua, Activo, APA, European Pressphoto Agency EPA and possibly more that we have missed!

One of the highlights for us was the Provenance and AI panel moderated by Anna Dickson, who until recently was with Google but is now with another IPTC member, Shutterstock. Presenters on this panel included Katharina Familia Almonte, product manager at Google, Andy Parsons, Director of the Content Authenticity Initiative  (although Brendan ended up presenting his slides due to it being 4.30am for Andy in New York at the time!); Mathieu Desoubeaux, CEO of IPTC startup member IMATAG, and Brendan Quinn, Managing Director of IPTC.

Brendan used the opportunity to introduce the IPTC Media Provenance Committee and the work that the IPTC is doing on creating a C2PA Trust List for media organisations. Brendan put out a call for other media organisations who may be interested in joining the next cohort of certificate holders who will be able to obtain a certificate stored on the trust list and use it to cryptographically sign their media content. The discussion went on to look at the issues around using both C2PA and watermarking technology to protect image content.

Later in the day was the panel “Where Law and Technology Meet.” Moderated by Lars Modie (CEPIC, previously of IBL Bildbyrå / TT in Swededn) with speakers: Serguei Fomin (IQPlug, IPTC CEPIC representative), Brendan Quinn (IPTC), Franck Bardol (University of Geneva), Nancy Wolff (partner at the intellectual property, media and entertainment law firm of Cowan, DeBaets, Abrahams & Sheppard, LLP and DMLA counsel) and Katherine Briggs of Australian agency Envato which was recently acquired by Shutterstock.

At this panel, Brendan outlined IPTC’s recent guidance on metadata for generative AI images. This includes the Digital Source Type property but also guidance on using the Creator, Contributor, and Data Mining properties to signal ownership and rights licensing information associated with images, particularly for engines that “scrape” web content to train generative AI models.

Many stimulating conversations always make the CEPIC Congress a valuable event for us to attend, and we are already looking forward to next year’s instalment.

 

… the image business in a changing environment

By Sarah Saunders

The web is a Wild West environment for images, with unauthorised uses on a massive scale, and a perception by many users that copyright is no longer relevant. So what is a Smart Photo in this environment? The IPTC Photo Metadata Conference 2018 addressed the challenges for the photo industry and looked at some of the solutions.

Isabel Doran, Chair of UK image Library association BAPLA kicked off  the conference with some hard facts. The use of images – our images – has created multibillion dollar industries for social media platforms and search engines, while revenues for the creative industry are diminishing in an alarming way. It has long been been said that creators are the last to benefit from use of their work; the reality now is that creators and their agents are in danger of being squeezed out altogether.

Take this real example of image use: An image library licenses an image of a home  interior to a company for use on their website. The image is right-click downloaded from the company’s site, and uploaded to a social media platform. From there it is picked up by a commercial blog which licenses the image to a US real estate newsfeed – without permission. Businesses make money from online advertising, but the image library and photographer receive nothing. The image is not credited and there is no link to the site that licensed the image legitimately, or to the supplier agency, or to the photographer.

Social media platforms encourage sharing and deep linking (where an image is shown through a link back to the social media platform where the image is posted, so is not strictly copied). Many users believe they  can use images found on the web for free in any way they choose. The link to the creator is lost, and infringements, where found, are hard to pursue with social media platforms.

Tracking and enforcement – a challenge

The standard procedure for tracking and enforcement involves upload of images to the site of a service provider, which maintains a ‘registry’ of identified images (often using invisible watermarks) and runs automated matches to images on the web to identify unauthorised uses. After licensed images have been identified, the image provider has to decide how to enforce their rights for unauthorised uses in what can only be called a hostile environment. How can the tracking and copyright enforcement processes be made affordable for challenged image businesses, and who is responsible for the cost?

The Copyright Hub was created by the UK Government and now creates enabling technologies to protect Copyright and encourage easier content licensing in the digital environment. Caroline Boyd from Copyright Hub demonstrated the use of the Hub copyright icon for online images. Using the icon (like this one ) promotes copyright awareness, and the user can click on the icon for more information  on image use and links back to the creator. Creating the icon involves adding a Hub Key to the image metadata. Abbie Enock, CEO of software company Capture and a board member of the Copyright Hub, showed how image management software can incorporate this process seamlessly into the workflow. The cost to the user should be minimal, depending on the software they are using.

Publishers can display the icon on images licensed for their web site, allowing users to find the creator without the involvement of – and risk to – the publisher.

Meanwhile, suppliers are working hard to create tracking and enforcement systems. We heard from Imatag, Copytrack, PIXRAY and Stockfood who produce solutions that include tracking and watermarking, legal enforcement and follow up.

Design follows devices

Images are increasingly viewed on phones and tablets as well as computers. Karl Csoknyay from Keystone-SDA spoke about responsive design and the challenges of designing interfaces for all environments. He argued that it is better to work  from simple to complex, starting with design for the smartphone interface, and offering the same (simple) feature set for all environments.

Smart search engines and smart photos

Use of images in search engines was one of the big topics of the day, with Google running its own workshop as well as appearing in the IPTC afternoon workshop along with the French search engine QWANT.

Image search engines ‘scrape’ images from web sites for use in their image searches and display them in preview sizes. Sharing is encouraged, and original links are soon lost as images pass from one web site to the next.

CEPIC has been in discussion with Google for some time, and some improvements have been made, with general copyright notices more prominently placed, but there is still a way to go. The IPTC conference and Google workshop were useful, with comments from the floor stressing the damage done to photo businesses by use of images in search engines.

Attendees asked if IPTC metadata could be picked up and displayed by search engines. We at IPTC know the technology is possible; so the issue is one of will. Google appears to be taking the issue seriously. By their own admission, it is now in their interest to do so.

Google uses imagery  to direct users to other non-image results, searching through  images rather than for images. Users searching for ‘best Indian restaurant’ for example are more likely to be attracted to click through by sumptuous images than by dry text. Google wants to ‘drive high quality traffic to the web ecosystem’ and visual search plays an important part in that. Their  aim is to operate in a ‘healthy image ecosystem’ which recognises the rights of creators. More dialogue is planned.

Search engines could drive the use of rights metadata

The fact that so few images on the web have embedded metadata (3% have copyright metadata according to a survey by Imatag) is sad but understandable. If search engines were to display the data, there is no doubt that creators and agents would press their software providers and customers to retain the data rather than stripping it, which again would encourage greater uptake. Professional photographers generally supply images with IPTC metadata; to strip or ignore copyright data of this kind is the greatest folly. Google, despite initial scepticism, has agreed to look at the possibilities offered by IPTC data, together with CEPIC and IPTC. That could represent a huge step forward for the industry.

As Isabel Doran pointed out, there is no one single solution which can stand on its own. For creators to benefit from their work, a network of affordable solutions needs to be built up; awareness of copyright needs support from governments and legal systems; social media platforms and search engines need to play their part in upholding rights.

Blueprints for the Smart Photo are out there; the Smart Photo will be easy to use and license, and will discourage freeloaders.  Now’s the time to push for change.