Not Just Nightmare People with Too Many Fingers
A few practical articles to put AI/ML in photography in context
Hello and thank you for subscribing to the Photo AI newsletter! Consider becoming a paid subscriber to support my work for as little as $4.17 per month. And for those wonderful souls who already support this endeavor financially, including the new subscribers since the last edition, an extra thank you!
David Etchells, writing at PetaPixel, sat down with Fujifilm executives to talk about their camera lineup in a discussion that touches on many AI/ML issues. It’s a good interview throughout, but what really caught my eye was a section about incorporating ML processing at the capture stage in mirrorless cameras. It’s something that smartphones do extensively, but is more challenging technically in other cameras.
I was also struck by the revelation that the autofocus speed was doubled in the Fujifilm X-H2S in a recent firmware update through optimizing the algorithms. That’s impressive. Read the interview here: Fujifilm Opens Up About AI, 8K Video, Entry-Level Cameras, and More.
A Pair of DPReview Articles from Me
I’m thrilled that not only has DPReview survived being killed off by its former corporate parent Amazon, it’s been purchased by Gear Patrol and will continue to live on. And that means I’ve been able to continue writing for this venerable publication.
Two AI/ML-focused articles of mine were published in the last couple of weeks:
I’m a keyword guy, I’ll admit, and anything that helps photographers tag their photos more successfully is a win. ON1 Photo Keyword AI scans a folder of photos (or scans them during the import process) and generates keywords based on what the app recognizes in them. That can be objects, numbers of people, image characteristics such as tone, even estimated ages of people that appear.
After scanning, the keywords are saved to the images (within JPEG files or in XMP sidecar files for raw images). If you’re using something like Lightroom Classic to manage your library, you can scan the images where they exist on disk and then update the metadata within Lightroom.
My chief criticism of the app is that it tends to generate too many keywords—I know, it’s shocking to see myself type that. But often there’s just too much detail: “human arm” is not something that’s going to come up in a search, for example. I find that Excire Foto does a better job at coming up with more targeted keyword, but it tends to create fewer of them. Still, having too many keywords isn’t a terrible problem to have.
I’ll let the opening paragraphs do the work for this description:
As tools based on machine learning and AI have appeared, most recently Adobe’s Generative Fill feature in Photoshop, photographers seem to bounce between embracing the technology as a new creative tool and rejecting the intrusion of “AI” into a pursuit that values image authenticity and real-world experience.
But while generative AI has stolen all the attention lately, machine learning has long maintained a foothold in the photography field. Here are five areas where you’re probably already benefiting from machine learning, even if you’re not aware of it.
A new episode of Photocombobulate is out!
Episode 31: Storm Chasing: Storm chasing has become a spectator sport in the United States. When colossal supercell thunder storms spin up in the U.S. Midwest, dozens of amateur and professional meteorologists tear after then in hopes of capturing images and video of intense lightning, hail, rain, and of course, tornadoes. Many of these storm chasers live-stream their pursuits on YouTube and other social media platforms, garnering some celebrity status with millions of viewers wing their exploits. Mason’s 12 year-old son, Cooper, has been closely following storm chasers and the severe storms they capture on video. These videos and the stories of the thousands of tornadoes that have ravaged parts of North America have become an obsession for him and he hopes to one day become a meteorologist who works to predict these storms and help people stay out of their way. Motivated by Cooper’s passion for severe weather, Mason promised him that if he got good grades this school year, they’d go storm chasing. This episode is the story of that adventure.
Here are links to the most recent episodes of my PhotoActive podcast:
Episode 146: Tea and Coffee: When we record PhotoActive, Kirk is at the end of his day in the U.K. and Jeff has just begun in Seattle, which means we both start off with steaming mugs of our favorite caffeinated beverages. For a change, we thought we’d make this the tea and coffee episode, essential ingredients for any photographer.
Episode 147: Is Photo Editing Too Difficult?: Kirk and I come out swingin’ at the start. Prompted by a listener question, we address the question of whether photo editing is too hard. There are lots and lots of options, and software certainly does amazing things, but is it beyond what most people really want to do? Is it distracting them from creating photos? Listen in and find out.
Generative AI for Designers, CreativePro Magazine (subscription required)
Thanks again for reading and recommending Photo AI to others who would be interested. Send any questions, tips, or suggestions for what you’d like to see covered at email@example.com. Are these emails too long? Too short? Let me know.