MicrostockGroup
Agency Based Discussion => Shutterstock.com => Topic started by: Sandeel on May 01, 2020, 03:44
-
What's up with these "similar content" rejections? I never had those, until recently. It's enough for a composite to have an element which I had used once before to get a rejection. In my case I find this extremely unfair, because the end results are not similar at all and I work for hours on every single image. At the same time I see whole batches of really similar content when I go through fresh content that they accept. Is there a way to make them see the injustice or is this a dumb robot deciding? :-\
-
Split up your similars into separate submissions.
That's what I do to avoid those rejections.
-
I think it’s a dumb robot deciding. But you are right, it doesn’t explain how some contributors can have multiple pages of almost identical images. I also think there is a bit of favoritism going on, i.e. approval of friend’s images. Who knows. And for me, it doesn’t matter anymore. They allowed content to be stolen by not closing up holes in their platform. I closed my account a couple of months ago, but likely too late.
-
i agree - I think quite possibly there is approval of friends items going through 100% (I think actually envato was doing that for quite some time, not sure if they still do, but if I recall - a number of the reviewers there were actually contributors as well)... while other items get 100% rejected right now... seems like they (the management) doesn't really care to monitor that right now...
-
Split up your similars into separate submissions.
That's what I do to avoid those rejections.
Thanks, but I have no similars in one batch. They reject images in which they find similarities to files that are several years old.
-
Split up your similars into separate submissions.
That's what I do to avoid those rejections.
Thanks, but I have no similars in one batch. They reject images in which they find similarities to files that are several years old.
Amazing isn't it. The machines look for patterns and colors. (sorry if it doesn't match any of the conspiracy theories) ;) I'm more surprised that they are going back years or months now, you must have some very distinctive elements.
I really think they have gone overboard on similar rejections. I can understand, the same shot, moved a little, and done 20-40 times, but I don't doubt that yours have one element of the composition that's getting a similar rejection now.
I don't know where that's leading, because instead of solving a problem, they just blew up everything with some dumb machine learning that's set up far to tight.
Just like the reviewers made mistakes and reject things that should be accepted, some will be lazy and accept images that should have been rejected. We're back to inconsistent reviews and standards. And that also explains whole batches for one reason, when that possibly only applies to one image. Gone crazy for fast and unreliable reviews. Well there you go, people asked for tighter standards, fast reviews, less duplicates and more rejecting of similar images. You got it.
Not what was expected and seriously flawed, but that's what happened.
-
same here image similar but not similar. I will try send in two batch
-
Is this consider similar?
https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111 (https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111)
https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664 (https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664)
by the way.. do you notice there are from different contributors...
-
Is this consider similar?
https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111 (https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111)
https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664 (https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664)
by the way.. do you notice there are from different contributors...
Yes they are the same, as in identical, and each has 100 images. Yes, this is why similar rejection have become so tight. New rules, there would be one image? :)
-
I think they created the problem for themselves with the accept-everything policy they had for a year or two.
If anyone remembers some of the highlights like the guy with 25,000 pictures of boats in the same marina and so on unchecked.
The brute force method of gaming the search.
Unfortunately they seem to have gone too far the other way now in trying to correct. No doubt theres AI involved (which as above isnt perfect, pattern/colour recognition amongst other things getting it wrong) and an element of lazy reviewing not actually checking.
To me its random, some batches i get no rejections, others i do, sometimes its most, sometimes none and so on. Theres no consistency which makes it hard to work out what they class as similar or not.
-
Is this consider similar?
https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111 (https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111)
https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664 (https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664)
by the way.. do you notice there are from different contributors...
Yes they are the same, as in identical, and each has 100 images. Yes, this is why similar rejection have become so tight. New rules, there would be one image? :)
Well, just only...99!
Wow!!! This is a waste of storage, bandwidth and human effort. Wouldn't a blank one "cover" every possible use by the user adding simple text? Is it so difficult even in windows Painting to add a word in an image!?
-
There is no Artificial Intelligence. Artificial Intelligence related to the choice of the image files has more value than google, Facebook and internet together. I have my theory about rejections, after much meditation, and I see it very clearly, the agency, they do what they have to do. I would do the same as them. I see it very clear now, finally. I cannot make it public because I am going to write an article about it.
-
Is this consider similar?
https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111 (https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111)
https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664 (https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664)
by the way.. do you notice there are from different contributors...
Yes they are the same, as in identical, and each has 100 images. Yes, this is why similar rejection have become so tight. New rules, there would be one image? :)
Well, just only...99!
Wow!!! This is a waste of storage, bandwidth and human effort. Wouldn't a blank one "cover" every possible use by the user adding simple text? Is it so difficult even in windows Painting to add a word in an image!?
People don't want to spend time to edit an image. They just want to quickly download and use it then move on.
-
Is this consider similar?
https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111 (https://www.shutterstock.com/image-illustration/file-folder-labeled-audit-multicolor-archive-325424111)
https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664 (https://www.shutterstock.com/image-illustration/privacy-concept-on-folder-register-multicolor-397698664)
by the way.. do you notice there are from different contributors...
Yes they are the same, as in identical, and each has 100 images. Yes, this is why similar rejection have become so tight. New rules, there would be one image? :)
that's funny... either one has stolen images from the other... OR, they are working together trying to 'spam' the engines...
they are different images - but one is based off of another (just replacing the text which is probably a 2 second thing) on one of the images...
-
It's funny looking at those 100 identical file folders...
A few years ago (2014 or 2015) I did a New Year's concept with signs and then the next year I submitted the signs blank as well as with the next year, both were rejected as being similar to the earlier one. It's the only time I bothered to object to a rejection and requested that they at least take the blank one but they would not.
And to my surprise the outdated new year's image still sells any time of year - I am sure it would do better if it was blank (up on other sites where it sells more often) but it shows what a waste all those 100 file folder images are - maybe 4 concepts terms and a blank one would be the right compromise? I've always thought that 5 similar but not identical was the limit.
In the old days when we could upload a set, there always seemed to be two out of the three or five that would do best at one site, and maybe two different ones would do best at another, which argues for giving some limited ability to upload a small set of similar but not identical images. It's often hard to choose the best one in a set and if you ask five different people you might get five different answers.
It shows they still haven't dealt with the thievery issue. Why isn't their AI searching the site for all those not just similar but identical images?