Spotting copyright violations

#1

Unfortunately as with many places on the internet there is an issue with users uploading copyrighted photos onto the site as part of their ‘observations’. Adding photos to which you do not have ownership is both banned by iNaturalist terms of use, and a violation of copyright law.

Photos identified as being a copyright violation should be flagged as such, which renders them invisible (except to curators) on the site. Flagging is done by choosing the ‘i’ icon at the bottom of the photo which brings up the photo details page, and towards the bottom of the page on the right is a link to flag the photo, in which copyright violation is listed.

How can you spot copyright violations ?

While there is no absolute method to spot them, there are a few tricks you can use.

  • if you suspect a photo may be downloaded and a violation, if you use Google Chrome as your browser, you can right click on the photo and choose ‘Search Google for image’. This will run the photo through Google’s search engine to see if it can be found, and if so, generate a list of links to other websites where the photo appears to be used

  • if you open the observation and the photo only takes up a small amount of the available space in the photo box, this is often a hint the photo is a thumbnail and has been downloaded. It is not a guarantee as the user may just have cropped their picture etc, but cases like this should be checked through the reverse search as noted above.

  • photos that show a watermark, such as ShutterStock or other commercial sites

  • obvious screenprints or captures especially when you can see elements of the search engine or other computer programs

  • very high quality photos submitted by users with low observation counts. Please note there are some extremely talented photographers among iNaturalist users, and of course they all start with a low number of records. Unfortunately too often these very high quality photos are simply taken from the web, but they should be checked carefully.

  • one of the simplest ways to find violations is that unfortunately users who do it once are prone to doing it multiple times, so if a user has been flagged as having uploaded a copyright violation, reviewing their other photos is never a bad idea.

  • clicking on the ‘i’ button to see the photo details page will reveal the meta data on the photo ,and often in areas like comment, copyright etc can reveal that the photo is sourced from elsewhere

  • records that appear to be illogical in time or place. For example someone submitting a photo of a Snowy Owl from Panama. Sometimes these of course are driven by computer vision suggestion errors, but these are good candidates to check

What to do if you find a copyright violation.

  • please note that photos that are in the public domain or licensed under Creative Commons etc are still not permitted to be loaded to iNaturalist as they do not represent the submitter’s experience of seeing the organism

  • while there is no official policy that I am aware of on this, it is my view that suspicion alone is not enough to flag a photo, you need to be able to document that it is in fact downloaded. If you suspect a photo may be a violation, but can not find it via a reverse search, rather than flagging, a polite,non-accusatory question on the record can be left on the observation

  • ideally if you find a photo is a violation, copy one of the hyperlinks to where it can be found on the web. After flagging the photo add a comment that the photo has been removed as a copyright violation and paste the link into the comment…

  • optionally you can also use DQA flags such as Date or location is inaccurate in addition to removing the photo, although removing the photo alone will still force the record into casual status.

  • if the violation is of a species that is not seen in the location where it has been submitted from (again for example a Snowy Owl in Panama), I personally will also reidentify the record to ‘Life’ to avoid any chance of it impacting range maps, checklists etc.

4 Likes
#2

Sometimes I screenshot videos I take to upload to iNat.

#3

Hopefully that is covered by the guidance that if you can’t prove it, don’t flag it.

1 Like
#4

I don’t like to change IDs on copyright problem observations, because it is in fact that organism, but I am always willing to give them every assistance in vanishing from the map. Because if they used a copied image, they can’t know the date or location, and the (now flagged) image is gone, so DQA becomes
Date is Accurate No
Location is Accurate No
Evidence of Organism No
there was a time when these were ID’d by others to Velociraptor, but that too is false.

#5

Changing the id of obviously bogus submissions has other benefits.

It only takes the user to vote yes the DQA to reset it to equilibrium, it takes more than 1 id to offset my explicit dissenting ID.

It removes it from any cases where people are reviewing records for issues, or even exports.

It removes it from any projects impacted by the fakery, unfortunately not the few grandfathered ones with auto aggregation, but most.

I am sure there are others.

1 Like
#6

good points Chris, I’ll add “Life” to my triple-tap

1 Like
#7

I literally just caught one for this exact reason. Photo was grainy and identified as something way out of range. Right click > Search Google for this image. Turns out it’s a stock photo of “grass” used on multiple websites.

I will say that I’m impressed how rarely I find copyright violations on iNat. Considering the volume of submissions, I’d expect them to be more common.

2 Likes
#8

Thanks for this great tutorial, @cmcheatle!

Just wanted to add that other browsers besides chrome have add-ons/extensions for searching Google images. I use https://addons.mozilla.org/en-US/firefox/addon/google-reverse-image-search/ for Firefox, but there are others.

3 Likes
#9

Just because a photo pops up in Google image search doesn’t mean it’s a copyright infringement. I noticed a new inat user had a couple high quality watermarked photos taken from a blog and after a little digging I found they write the blog and they had taken the photos.

4 Likes
#10

Regarding “photos of photos/screens”, rather than “signs they are infringing”, i think of them as triggers to ask some probing (but polite) questions… usually framed in terms of “i have marked this as ‘no evidence’, let me know if you actually saw it”… I know I’m active and will be around to change my DQA, but too often the observer never responds…

1 Like
#11

I really appreciate getting to look at so many excellent beautiful photos on the site from some very talented photographers. When I see a really great photo from a user I don’t recognize I click on it to see their other photos too. Sometimes then I see the clues that were already mentioned that lead me to do a reverse image search. If the photo is linked to on blogs dated years ago, but the user states they took the photo the day before that is a pretty good clue, but I still try to phrase my comment in a not so accusatory fashion. I think it’s possible the observation date was taken from the file date which may have been incorrect and the user just didn’t fix the observation date afterwards.

In about half of the cases that I’ve seen though, the user actually points out in their description that the photo is not theirs, but it “looks just like” what they saw. Quite often those observations have even already reached research grade. Many identifiers must not read observation descriptions very often.

1 Like
#12

There are a lot of identifiers that work off the identify page and use the agree button, so they don’t see descriptions or even comments on the observations. We need to lose those “Agree” buttons…

3 Likes
#13

Would that be appropriate for this case here? https://www.inaturalist.org/photos/35220647

1 Like
#14

In this instance the observation was uploaded using the Android app. A small sample of her observations suggests she is only using the Android app and thus would not be seeing comments. The user is in the habit of taking images of screens, at least screens of a camera:
https://www.inaturalist.org/observations/18338483
https://www.inaturalist.org/observations/17579901
https://www.inaturalist.org/observations/17579379
https://www.inaturalist.org/observations/16979999
This appears to be her way of uploading images captured with her camera: take a picture of the camera screen with a smartphone. Given that the images are in a camera, the benefit of the doubt has to go to those being her images. Her workflow for getting images into iNaturalist is a tad different. The location may be problematic: a number of the locations appear to be the location at which the camera was imaged by the phone which appears to have occurred in a residence.

1 Like
#15

And as observation photos are published on iNat credited to the observer, with their copyright, this is copy-fraud anyway - so is indeed a form of copyright violation.

1 Like
#16

This is quite common, some find it easier to do than downloading photos from their camera and uploading them to the website (or perhaps don’t know about the website - another topic!). Some also find it convenient to post observations from the field by using the app to capture their cameras preview screen, then upload better quality image when they get back to base (especially if their camera doesn’t have GPS).

3 Likes
#17

Another good resource, that has been around longer than Google’s reverse image search, is https://tineye.com/. It also allows you to sort the results by date and hence find the earliest, and hence possibly original, posting of the image.

2 Likes
#18

Comments are on there, you just have to drop down the menu to get them. Which you are right, she may not notice.

2 Likes