Continuing the discussion from Suggestion on how to include habitat images:
Ìs this written somewhere in the help section? I couldn’t find it. It this is the rule, it would be good to have a way to link to it. (More official than a forum post.)
1 Like
https://www.inaturalist.org/pages/help#observations1
Photos attached to observations should include evidence of the actual organism at the time of the observation, observed by the user who is uploading the observation.
4 Likes
Thanks, I missed that, not sure how. I wonder though if people might think ‘yeah, my first 2 photos do show the organsim, so the photos I have attached do include what is asked for here, regardless of what my other photos show’.
2 Likes
I think a lot of us do more loosely take that approach. I’ll include habitat shots in which the organism may not be visible (it may be in there, just too small to discern) sometimes. If you want a broader “habitat shot” for a nearly-microscopic liverwort or something then there’s really no way around that. It’s sort of like the rule that observations should be of a single discrete organism- a lot of us will use an observation to represent a whole population (say, a rare plant with just a couple dozen plants present) with photos of a few representatives and observation fields with additional info. I’d argue that flexing these soft-rules in cases like these creates more useful data, and both practices are pretty common among well-informed users.
2 Likes
There’re many ways around and habitat shots shouldn’t be included, add it in a comment, present a link to another site, etc. Photos of grass shouldn’t be there if you search for an insect on its page.
5 Likes
At the conference I went to this past weekend, there were several posters/abstracts of student projects mining iNat observations for habitat information to answer a scientific question. Along with looking for species interactions, this was probably the most common type of “iNat research” presented besides looking at ranges and phenology. So this data is valuable for current research and making it easier to incorporate and search for could be of benefit for the scientific community. I wished there was a way to tag/annotate individual images and not just the observation as a whole - that might be a way to get at this.
The main problem I see with asking people to put such pictures into comments is that 1) not everyone has an external account to put those pictures, and 2) a lot of those other websites have proven unreliable in terms of longevity, eventually leaving broken image icons in place of the pictures that were originally there. Here in the forum, if you link to an external image it gets downloaded onto the server to preserve it for the context of the thread it was linked in. I don’t think that happens with images in the comments on iNat.
4 Likes
Your photos will still be with you, if link is broken after three years, just reupload it. All that info can be presented via observation fields.
2 Likes
My own images, yes, I keep everything. Other people’s images, no. I’ve come across plenty of forum threads on other websites looking for specific info that used to be there but are now full of broken image links.
3 Likes
I didn’t realise you can insert pictures into comments. How do you do it? I have just tried the usual cut and paste, control-C and control-V, but that didn’t work.
1 Like
No, she meant a LINK to the picture, not the picture itself. The link would take the person interested to a place where you have stored that image, say your Dropbox or Instagram account for instance.
2 Likes
https://forum.inaturalist.org/t/useful-html-tags-for-inaturalist-comments-and-other-text-wiki/6198
I meant a photo, via link. Like here https://www.inaturalist.org/observations/39751671
3 Likes
I disagree. The environment is part and parcel of the information contained in the observation. Not a landscape shot just for the heck of it, of course, but to give an idea of the size/place/habit and habitat of your subject (note: habitat where your observation was made, ie the subject is in there in the picture, small perhaps, but there all right). OOne picture of, say, the inflorescence, another of the entire plant which gives an idea of where you’ve seen it.
3 Likes
Well, if subject is in the picture, small or big, it’s legit, it’s not what we discuss here, usually people post generic landscape shot as an addition, often one shot for many observations.
4 Likes
Perhaps you could provide a how-to to @jhbratton how you achieved that comment, ie what did you type into the comment box to make a photo stored on an another server appear inside the comment?
When I link to my Dropbox account, the media are rendered by Dropbox in another browser tab, not within iNat.
I think the most debate happens when the observation itself is a sound file, accompanied by a habitat shot showing where the sound was heard but not a picture of the organism that made it. Also, often the argument is made that such pictures may confuse the CV algorithm. Hence, it would be good to have a way to include them but without being picked up for the CV training set.
1 Like
I see the point with regard to the CV algorithm, especially with sound files.
And I agree with you that it would be ideal if we could exclude such environmental shots from the CV’s to-do list, ie have them ignored by the algorithm.
Does anyone know how things are handled in the case of bat sounds? Example: https://www.inaturalist.org/observations/152396588
There is the sound and the graphic representation of it. What does the CV algorithm do in such cases?
That’s exactly what I did? Check the topic I linked.
Thanks. I don’t understand it all but am I right in thinking you can only put the photo in a comment if it is already on another website? Mine are just on the laptop and a portable hard drive.
You can use sites like flickr, photo storages or just pm it somebody on a social media, any link will do.
1 Like
Putting habitat photos and whatnot in links in comments is a bit of a kludge- the “rules” should adjust to behavior the user base has shown to be desirable (and as Annkatrinrose notes has been useful for research), rather than insisting on elaborate workarounds requiring that users add entire separate image hosting services to their workflow.
Habitat is useful visual information and those photos are better included in the observation itself. The only issue with that is CV training, and that would be better solved by having a way to tag images on upload to be excluded from CV training.
4 Likes