USING AI TOOLS TO ENHANCE PHOTO DISPLAY
AI, a gift from God or a toolbox from the devil?
It’s both.
By Ray Saint Germain • Bay City News
I have been a photojournalist covering the Bay Area for over 30 years and in the days of film and darkrooms we followed a code of ethics.
We didn’t stage photos or allow subjects to set them up.
We showed our subjects respect and did our best to truthfully portray them.
We strived to cover all sides of an issue, allowing our readers to decide what side they wanted to be on.
It was a simpler time – we would shoot an assignment, grab a coffee when the film was processing, invite an editor or other photographers over to the light table to make some picks, print in a darkroom, put a caption on the photo and then send to the desk for publication.
Post-production manipulation wasn’t a thing.
In the early 90s, darkrooms were closed and computers loaded with Adobe Photoshop became the norm in newsrooms for post-production and the ability to manipulate an image became much easier.
Realizing this, the National Press Photographers Association adopted the NPPA Code of Ethics.
The game changed again in 2022 with artificial intelligence image generation.
As described by Adobe, “Add, remove, or expand content in images right in Adobe Photoshop with simple text prompts powered by Adobe Firefly generative AI.”
As a photojournalist, generative AI is definitely a tool of the devil, let’s call it ‘Satan’s spanner’ wrench since it can take things out of and add things to an image.
Photojournalists have a responsibility to the public to be honest and objective while on assignment and if we abide by the NPPA Code of Ethics we’ll be fine.
I mentioned one of the devil’s tools so I must mention a few of God’s gifts.
Time-saving AI tools such as noise reduction and increasing resolution, both shown, are ethical uses of AI use in editorial photography.
Denoise:
Gigapixel:
Also, Photoshop’s selection tools are now much more accurate and efficient using AI.
Evolved from the Lasso and Magic Wand selection tools, most often used for color correction and tonal changes, the Object Selection, Quick Selection, Select Subject and Select Sky tools are now much more accurate and efficient.
As an industry we are in the early stages of developing hardware and software tools to counter the increase of manipulated photos being published and posted.
These tools show the provenance and post-production changes made to an image.
Bay City News partnered with The Starling Lab for Data Integrity at Stanford University and the University of Southern California to advance authentication tools for photojournalism.
Using authentication tools, a specialized “HTC crypto phone,” Starling’s phone app, Canon R5 cameras, Adobe Content Credentials and the Four Corners Project, Bay City News documented the Stockton homeless crisis.
Bay City News’ photo contribution to the project was close to 100 photos signed by Starling Labs over months of covering Stockton’s homeless population. During this time, photographer Harika Maddala also covered general assignments and news stories.
Fifteen stories were published through the course of the Starling Lab project and are all presented on a LocalNewsMatter.org project page with 35 authenticated photos and six authenticated photos displayed using Four Corners.
This project was important because now most news that is relevant and actionable to readers is local and we know that trust is an issue given the misinformation and disinformation that proliferates. This is true in Stockton, for example, with the upstart 209 Times website and other outlets not based in traditional journalistic practice, the ubiquitous Photoshopping on social media platforms where many get their news and reduced coverage of local news published by trusted sources.
Being able to prove the origin and authenticity of photos could be useful to professional newsrooms like ours as well as the public. It is imperative that local publications reassure readers that our facts are verified and our images are real. The advent of new cryptographic and decentralized tools that facilitate the capture, storage and verification of time, location, device and photographer now being engineered into professional cameras, will help us reach that goal.
Through the Content Authenticity Initiative (CAI), a group driving the implementation of an open industry standard for content authenticity and provenance, and the Adobe-led Coalition for Content Provenance and Authenticity (C2PA) that is working to establish technical standards for source certification and provenance of media content, the tools and methods of authentication are making the process easier.