Copy link to clipboard
Copied
I use MidJourney to create all of my images. To date, MidJourney has been able to produce fantastic images in a photographic style. It seems that every time I upload a photo-style image, it triggers an unexpected response. I made sure to check off the appropriate boxes indicating that I was submitting an AI-generated image, yet I receive follow-up questions regarding the authenticity of my upload. This left me perplexed with certain uncertainties.
Since I am the creator of the art using MidJourney, do I still need to submit a release form? Is my checking off the boxes indicating that the art I'm submitting was created by me and is equivalent to me stating, "Hey, I'm releasing the AI-generated art I created"? Or am I completely wrong in that assumption?
[Moderator moved the thread to the correct forum]
If your AI images are labeled as photos, even with all the correct boxes checked, you will still get a reminder that some of your images may require a release. This is apparently just a reminder to double check everything and you can continue to submit the images for review. To elimate the issue entirely, label the images as illustrations (which I do, if the images are more painterly or illustrative stylistically than actual "photography." But it has been noted here that Photos can indeed be sel
...Copy link to clipboard
Copied
If your AI images are labeled as photos, even with all the correct boxes checked, you will still get a reminder that some of your images may require a release. This is apparently just a reminder to double check everything and you can continue to submit the images for review. To elimate the issue entirely, label the images as illustrations (which I do, if the images are more painterly or illustrative stylistically than actual "photography." But it has been noted here that Photos can indeed be selected for photo realistic images.
Copy link to clipboard
Copied
I thought about doing that -- labeling each image Illustration. But some images I create in MidJourney look so photo realistic I struggle with calling them illustrations. Oh well...I guess I have to live with the situation as is. Thanks for your feedback.
Copy link to clipboard
Copied
Hello Steveeyes,
Your experience with MidJourney's AI-generated images and the subsequent inquiries about their authenticity are indeed intriguing. Despite your efforts to label your images correctly, these questions persist, leading to uncertainties.
It's noteworthy that even if certain MidJourney-created images possess an exceptionally lifelike quality, some complexities are at play. If you're determined to ensure the acceptance of your generative AI-generated content into the Adobe Stock collection, I'd like to furnish you with a comprehensive guide for submitting authentic assets that align with the platform's quality, legal, and technical criteria.
Prior to publishing on Adobe Stock, I recommend a thorough review Adobe Stock Generative AI Content Guidelines available here at https://helpx.adobe.com/stock/contributor/help/generative-ai-content
Feel free to explore these insights and the linked guidelines for a more informed perspective.
Copy link to clipboard
Copied
I appreciate the input and the link. I read everything I could find to read, including your link. My ex military organization jumped in, and I printed everything out, put it in a binder, read it, highlighted it, and so on. I compare my inputs to inputs approved by Adobe and I must admit, I think most of my images look as good if not better. It adds to my confusion. There is one thing that may be in play, and that is subject. Anyway, I love doing the whole Midjourney thing and I'm happy even if some of my images don't make the grade.
I wish there was a tool I could input an upscaled image, and it gets graded or evaluated in some way. I'm searching but can't find such a tool.
Copy link to clipboard
Copied
It seems that Adobe has a pretty high bar for acceptance of "photo-realistic" AI assets in the Photos file type. I've seen some images here and on the Discord channel that the Contributors are convinced are very realistic, and I can spot in an instant the signs that they're really AI, and not all that realistic. I think if I were you, I would continue to submit them as Illustrations.
Copy link to clipboard
Copied
The reminder screen is new. And they did probably forget to incorporate the generative AI rules (and btw, the Illustrative Editorial content). Looks more like a makeshift warning, as Adobe seems to be aggressively locking down on accounts that do not follow any rules.
I would, for generative AI, ignore the missing release information, when the “fictional character or property” flag is set. That is equivalent to a release, and simplifies the moderation (and submission) a lot.
To be clear: Photorealistic assets can be submitted as photos. If the assets are not photorealistic, submit as illustration.
Copy link to clipboard
Copied
OK, let's play a game. One of these four images was a photograph I took about 26 years ago. She was my workstudy at a college where I used to teach. The other three were created by Midjourney useing the /describe function, where I take one of my own images and then let Midjourney create prompts based on the image I submited. After removing any references to other artists or photographers (apparently, my own work was inspired by present day artists or photographers 26 years later), I selected one of four prompts and let Midjourney do it's thing. Can you guess which one was the actual photograph? Note: all four images were edited to one degree or another. You'll probably be able to figure it out, but this is to show just how good AI is getting. And it's going to get even better.
And I also believe that people who are turning off AI in their search function are doing themselves a disfavor, if not now, then eventually.
Copy link to clipboard
Copied
You have responded to my contribution; however, it appears that you are not specifically targeting me in this instance. 😂
Copy link to clipboard
Copied
No...I was not. These threads are hard to follow by the third glass of wine. My bad. 🙂
Copy link to clipboard
Copied
No...I was not. These threads are hard to follow by the third glass of wine. My bad. 🙂
By @daniellei4510
Every so often, it helps to use linear view against threaded view and vice versa.
😂