Unclear reasons: Similar content in our collection?

Nancy, you are kidding.

As I already said I am 1000% sure that in this case there are no pictures from that location in the database! So I did my job clearly.

The hints about keywords can be helpful because many of them are often generic like lake, bavaria, spring, and only one very specific like Großhöhenrain.

However, I think if there are strong overlaps regarding keywords and the pictures are much different, e.g   regarding length, width, focal length, colors, etc. They are often very different.

HOWEVER, in any case it would be SO much better to get detailed feedback about REASONS for similarities:

1. Adobe to show us 2 pictures regarded as too similar 

2. Which keywords are leading to duplicates 

3. Whatever other reasons leading to similarities 

For a programmer implementing this is a piece of cake, because it is actually already programmed, just an output of the reasons has to be made.

My observation is that rejection due to similarities is now running different compared to the past, and it might be somewhat broken even, i.e. rejecting accidentally far too early. But as the reasons are not named in detail we do not know!

Algorithm Flagging: Adobe Stock uses a shallow AI system to compare your submissions to existing images, vectors, or videos. If your work is too close in composition, concept, or subject to assets already in the library, it may get flagged for “similar content.” 

Did you compare your content with current Stock inventory? 

How much competing content was there— a few hundred, a few thousand or several million assets?

Repetitive Metadata: Using same titles and keywords across multiple variations can trigger automatic rejections.  Wherever possible, use brief but relevant distinctive titles & keywords. The less fuel you throw on the fire, the better. 

More about “Similar Content”

https://helpx.adobe.com/stock/contributor/help/how-to-submit-distinct-content.html

Hope that helps.

As ​@yamato713108855 indicated, it is overused keywords and titles that result in an asset being rejected as similar content. Unfortunately, these overused keywords are compared to the entire database not the contributors alone. It’s a less that perfect method in my opinion, but one contributors have had to live with for at year the past year. I maintain, perhaps incorrectly, that the issue is not only keywords themselves, but the order in which they are placed. It is not the moderators making this determination, but an algorithm. Assets are reviewed by human moderators first. If they pass quality, they are then exampled by an algorithm to determine if they are overrepresented. This can determined in cases where an asset is rejected on quality issues first, corrected and re-submitted, only to be rejected for similar content.

I maintain a three-column spreadsheet of all my submitted assets containing its general description or file name, the title as submitted, and the keywords used for that asset in the order as originally assigned. After a few weeks or even months, if I firmly believe that the asset is in fact unique, I consider resubmitting it with more unique but still accurate keywords in a revised order. (As someone who submits AI, I also have the advantage of making slight or even major changes to the original prior to resubmission). I don’t do this on a regular basis, but it has worked to my advantage on a few assets.

All that said, there are occasions where assets are rightfully rejected as similar. The database has more than its fair share of cats and dogs, wallpapers, flowers, icons, etc., that do indeed warrant a similar content rejections. But even those will be accepted if they are presented a unique and captivating manner.

Hi Ralph, the reason for my post is that Adobe seems never to give us clear reasons for rejection due to similarity. As said, it would be nice to get e.g. 2 examples from the existing content or e.g. we have already 100 pictures from the White House.