Management of pictures of specimen "artistically" altered

Hello everybody
I sometimes spot pictures of Fungi altered with scribbles or drawings (example here https://www.inaturalist.org/observations/27613779) - which, even if I can feel some simpathy for the ‘artist’, I find not in line with the goal of providing a gallery representing the species.
When I browse the gallery to try to solve a difficult identification, pictures with smileys or “I love x” are not of particular interest to me and I think many other people trying hard to keep the platform as scientific as possible. My opinion maybe, but my understanding is that iNat hasn’t got the same goal as other picture sharing platforms.
I wonder if there is any policy on this (the risk being obviously an accumulation of spoof pictures) - I tried flagging for revision but I got back a plain “not an appropriate reason to flag”

Any ideas?
Thanks

2 Likes

While the examples you give are not necessarily welcome, I can think of one time where altering a photo could be beneficial. This would be when someone adds line with text pointing out identifying features as seen in many field guides. Not that many would be interested in doing this.

1 Like

Agree, absolutely.
But that’s different. I mean, drawings or alterations not aimed at pointing at a feature of the specimen are prone to a questionable range of phenomena, from vulgarity to politics etc.
So if someone takes a pictures of - let’s say - a tree trunk on which he writes some obscenity and loads it as a sample of that tree - there is no mechanism to clean that up?
Sounds like a surprising loophole to me.
At least I think a mechanism should be in place to exclude the picture from catalogue and - even more - by the machine learning database used by inat.

1 Like

i feel like i came across a kind of similar, but not the same, issue the other day. filters.

like this one:
https://www.inaturalist.org/observations/36277758
for the sake of making it artsy, now its way harder to identify because the colors are gone

2 Likes

in this specific case, isn’t the ability to scratch out a picture on the fungus one of the distinguishing characteristics of the artist bracket fungus? doesn’t a demonstration of that characteristic actually help in identification?

11 Likes

https://www.inaturalist.org/observations/27613779

In that specific case, it also gave me an idea of what the name meant, what the fungus looks like, (which I had not learned from other observations casually encountered, but recognised in this obs).

And as it is actually artistic it felt kind of appreciative of the plant, so felt ok to me.

But in general it does sound horrific, esp. writing.

In the case of vulgarity or politics, flagging would presumably see it removed.

3 Likes

iNat’s primary goal isn’t to provide a gallery representing a species, although that is a byproduct of the observations shared on the site, and a valuable one. It’s to help people engage with nature. As far as I can tell there is nothing wrong with this photo, it’s evidence of an encounter with an organism, which is what an iNaturalist observation is. The reason it comes up first in the taxon photo browser for Ganoderma applanatum is because it’s been faved the most.

Are these common sights on iNat? The vast, vast, majority of the photos on iNat that I see are in situ photos of organisms, without any writing or other alterations.

And yes, if there are vulgarities, etc., those can be flagged.

8 Likes

That’s the example of what MANY people do with ganoderms, I see no problem with it other than destroying the fruit body, I’m not a big fan of that for species easily ided without doing so.

2 Likes

all right I get the point that the approach is more generalist.
Still I believe there shoud be some option in place to filter “artistic” content out - There are people not interested in that (or, viceversa, there might be people interested only in artistic content “connecting people to nature”)
Still personally believe that a carved wooden board of let’s say a pine tree is not a suitable observation for the tree.

2 Likes

yep, but, I mean - trees are used as a lumber source, but I don’t think posting a picture of a wooden hut should be an appropriate observation for the tree. But I probably got this wrong.

If the date and place is correct (e.g. photo of the plank that was made from the exact tree or tree population) it will be appropriate, otherwise it will be casual.

2 Likes

as long as it’s done in a sustainable way and in a generally tasteful (not vulgar) way, i don’t see any problem with a carved wooden board of a pine tree as one piece of evidence of the organism. i wouldn’t encourage chopping down trees just to get an observation like that, but suppose a big tree was going to be taken down as part of construction. if you got a photo of the tree before and after, and you were also able to get a section of the tree trunk, and if you took that tree trunk section and turned it into a nice table or some work of art that showed off how many rings were in that big tree… i think that would be a nice way to observe and memorialize that tree (though it would be sad that the big tree was taken down).

maybe the concern is that the fungus is destroyed as part of the observation. my impression from the mushroom comments on this forum is that taking the fruiting body is generally not a problem for the organism, since the fungus will still exist in its substrate. it’s true that the next person will not be able to see the fruiting body after it’s gone, but a new one should emerge in time.

in a way, i think the revelation of interesting characteristics of various organisms may help some to appreciate nature even more. that’s not just a nondescript fungus. that’s a fungus that artists have used over the centuries to make interesting art. and if that helps to engage people with nature more than a simple picture of the fungus on the tree trunk, then the artist bracket observation seems to be ok to me.

i think this is a slippery slope. if you create a filter specifically for “artisitic” content, then why shouldn’t there be a filter for animals that have been tagged or banded, or why not a filter for insects that have been caught in a net or a light trap, or why not a filter for plants whose flowers have been plucked and dissected to show the interior structures? any of those activities executed badly could be a model for bad behavior, but done well can help advance our collective knowledge and appreciation of nature.

you can already create an observation field to note “artistic” content or put such observations in an “artistic” project, if you wanted to. from there, you could filter out those observations based on their inclusion in that project. and you can flag bad observations outright. i think those existing tools provide you the options you’re looking for.

4 Likes

How is the computer vision/machine learning of iNat affected by artistically altered organisms?

1 Like

i don’t know how the computer vision works exactly, but i’d be surprised if an artistically altered observation threw it off any more than human fingers in a photo or other organisms in the background.

2 Likes

Almost minimally I should expect. There are over 6,000 research grade records for the species, which means at a minimum that number of photos.

3 Likes

Well, I feel that there should be a more structured way to filter out significally altered content, such as painted animals, painted insects, pictures with colors clearly altered and the like.
I am sure that, like me, there a significant amount of people who share a different sensitivity to “artistic” alterations (comes to mind my fellow people belonging to Insect amateurs here in Italy, UK or Austria/Germany, whose opinion I know for being a member of such groups).
I feel would be a sensible way to avoid putting off people with a more scientific approach/sensitivity towards alterations and let them contribute without having to deal with what they might consider clutter.
I am thinking of a feature request, where a user can opt in or out to see pictures flagged as altered.
I don’t think that would harm the artisty nature lovers at all.

it obviously depends on the species. Some of them do not have a huge amount of pictures.

In order to qualify to be included in the computer vision tool, a species requires a minimum of 100 photos, thus even the highest impact any such photo could have is 1%, and that is highly unlikely.

4 Likes

Thanks, this is important news. It also means that AI is not helping to identify species with a low number of observations. Good to know!

There are also pictures “artistically” altered.

See the 2nd, 3rd and 4th pictures in this observation:
https://www.inaturalist.org/observations/14628108

I guess the intention was to blur the background, to better see the foreground?