Computer vision/suggestion. Sorry I didn’t clarify. I see it worded as CV more than the actual words so I did that too.
2 Likes
I do this a lot too. Sometimes it’s way off for one but spot on for another.
2 Likes
I use it mostly to save me the work of typing the name in.
Sometimes I use it if I know the organism, but can’t remember the name.
And sometimes I use it if I have very little idea what I am looking at. Often it helps jog my memory or helps narrow down my options a bit.
11 Likes
To fill in ids I’m sure in, to look if it recognizes the object, so I look up the ancestors of it, it’d be bold to just use the suggestions it gives, I spend each day by looking at big amounts of horrendous ids proposed by cv.
6 Likes
I use it much as others have described. However, for Australian moths it has been a great help. In the past I would page through the plates in Ian Common’s “Australian Moths” or go through page after page of some of the websites to find what I was trying to id. We’re lucky that there has been a group of prolific and expert Aussie mothers who have greatly improved the CV. I now use it all the time with moths - to confirm an Id I am fairly sure of, or to get me to a family or genus that I can follow up. I don’t use it blindly but it certainly helps to get me on the right track.
11 Likes
I mostly do arthropods, and while it has gotten better at Australian species recently, I still mostly use it to as a guide to narrow down to family (except when I’m tired and cbf, then I just follow the suggestion if it looks right, but usually only as far as genus: iNaturalist is waaaaaay too confident on its species level IDs, especially on species that can’t be keyed from photos).
3 Likes
I think I’ve used it in most of the ways above at some point or another. Looking at the question ‘how I use the CV’ from a different angle: when identifying I use the symbol that tells you whether someone has used the CV to decide whether to leave an explanation when disagreeing. If they have not used the CV I will presume that they have put some effort into trying to identify it themselves, and so I will be more likely to explain why I disagree; if they have used CV I will guess that they probably just want to know what the thing is, and are less invested in their own ID - so I will rarely give an explanation unless they subsequently ask. It’s not a judgement of the user’s experience or knowledge, but of their investment in that particular ID - I know that it will sometimes be wrong and they’ve just used a shortcut, but it cuts the workload significantly and people can always ask.
12 Likes
@matthewvosper Welcome to the forum!
Most folks have covered how I use it. To save time. To narrow down choices ( I have the same experience as @nyoni-pete, made worse by the odd fact that Moth Photographers Group takes forever to load each page). To jog my memory. I rarely use it to identify to species - with moths there is so much variation that it’s best to take a look at as many as possible. So, I find it useful, but don’t rely on it except to save typing time!
1 Like
Computer vision 
.
2 Likes
On my own observations:
- if I know what it is as a shortcut, as others have mentioned
- if I don’t know what it is for a quick look at what it might be out of curiosity - after that I will add the ID I am sure about, often on class level (yep, I am that ignorant in most taxa)
On observations of others as a tool for narrowing it down. I remember one case where someone flagged me for help (I’ve made so many easy IDs that I turn up in IDer leaderboards on taxa where I don’t know anything about 95% of the sub-taxa in there) and I had no clue whatsoever. CV quickly showed me the right genus and the suggestion tool based on observations in the region quickly gave me a very good candidate which I could offer in a comment. If I recall correctly it turned out to be correct in the end (as per the IDs of other IDers). I totally love the toolset iNat gives us, of which the CV is one important ingredient.
8 Likes
It depends.
When I’m working on unknowns in India for instance, I like to see what computer vision comes up with to confirm or contradict my hunch. Sometimes the results are hilarious, sometimes informative.
When I’m working with plant, animals, etc. in the Carolinas, it can sometimes save typing, but not always. The CV struggles to identify fungus properly at times, but that is hardly surprising. However, over time, I’ve seen it get better at recognizing certain pupae as Campopleginae rather than bird droppings so that is encouraging.
I remember when software programs to correct grammar first came out, my English professor was horrified. She thought students would stop studying grammar since they could turn in correct papers effortlessly. I knew her fears were unfounded, but she didn’t believe me. So the next time I turned in a couple of paragraphs over that week’s reading I used my grammar program uncritically. I took absolutely every single suggestion it made. To give proper credit the last sentence I wrote was “Grammatik rules!”
She said I had written nonsense. I told her not to blame me. It was the grammar program not me.
That was early days then. As we know now, there are no more English professors or professional writers left. We all write equally well with auto-correct.
I like to think (and
the sooner the better!)
of a cybernetic meadow
where mammals and computers
live together in mutually
programming harmony
like pure water
touching clear sky.
I like to think
(right now please!)
of a cybernetic forest
filled with pines and electronics
where deer stroll peacefully
past computers
as if they were flowers
with spinning blossoms.
I like to think
(it has to be!)
of a cybernetic ecology
where we are free of our labors
and joined back to nature,
returned to our mammal
brothers and sisters,
and all watched over
by machines of loving grace.
-Richard Brautigan-
9 Likes
Some people? Think we all have a row of monographs and time to work thru keys and descriptions.Some people who live in ivory towers?
But even some highly skilled biologists use CV as a tool.
6 Likes
I read that they’re retraining it to indicate it isn’t as confident as it seems it is, so hopefully that limits the possible misuse of it. When I say misuse, I really only mean using it every single observation and then not ever engaging in conversation with other IDers. That’s the main misuse I think that exists, other than using it and hitting deliberately wrong ones.
4 Likes
Right there with you. Thought it was sort of a hurtful indication to make when I read that initial comment about it. Didn’t like it at all.
2 Likes
I’m glad you said this! When I go through the “life” and “unknown” listed observations, many of them are people who used the CV and it was super wrong, and somebody tried correcting them without leaving any type of comment about why they’re doing that. A lot of these observations come from people who don’t have many, so I don’t think they really have the hang of things. The use of the CV makes it easier to sort of nudge them in the right direction by explaining that the CV can be wrong and makes certain mistakes serially. That’s taking more time and effort, especially when majority of them never answer, but I’ve seen quite a few people actually change the observation to a vaguer ID, agree with somebody else’s or make a different observation without trusting the CV.
3 Likes
My brother-in-law’s house has a moth infestation on their ferns every year. I know what the moths are, but according to the CV, they are every other type of moth under the sun except the one that I know it is.
I mention that just because I know there are many specific animals it messes it up a lot, and moths certainly are one of them.
3 Likes
I use CV the way most people use. So I use CV as an alternative to typing in the ID but mostly I add broad ID so it comes faster when typing than through CV because typing d or f etc gives me what I need like dicots, ferns, fungi etc. Also I use CV to get a good guess and take it and go up the taxonomy and add an ID. CV gives ridiculous suggestions many times. I also do due to curiosity because I’m pretty sure it won’t give correct ID for many of my pics.
2 Likes
I’ll use the CV to try and help pin something down that I don’t know, especially since I am in the US which, yknow, pretty well sampled place. If the organism obviously isn’t anything the CV is putting up (the other day it thought my neonate redback salamander in northern new jersey was a scorpion, when the nearest ones are bark scorpions in Virginia, I think) I’ll go to the broadest thing it probably is, but if it looks good enough I might put down the CV id
2 Likes
I’m actually impressed with how good the CV seems with bees. I’ve learned a lot about identifying different types pretty much solely from the CV. I’d imagine since many look so similar that it would have a hard time differentiating but I think it does a pretty job with them.
4 Likes
Saved me some typing because the above is pretty much my exact usage as well. When I first started using iNat I thought I had to research in depth every single thing before uploading to try and get the best id possible and you know… not look too stupid!
Then I started having more days where I was taking tons of photos that were just piling up. I found that I don’t do well with having a big backlog of old photos to research and upload. I do better when I upload them within a few days of taking them, when the observation is still fresh in my mind. That way I can add notes to the observation and if someone asks me about some detail not visible in the photo, my memory is better. Also if they say I need a different angle for id, sometimes the organism is still available to go back and take more photos. I also like viewing other peoples observations in my area in close to real time. Seeing that a cool bird was in the area three months ago isn’t as much fun as seeing it was nearby yesterday and might still be there! Having the CV to help with quick uploads and rough ids gets those observations added more quickly, and I can always go back and refine or fix them later.
I also figure that even seeing or picking a wrong CV suggestion may be helpful. If it looks similar to another species, that gives me a starting point. I can check the “similar species” tab on iNat to see look-alikes commonly misidentified. I can check BugGuide under their “see also” section for similar looking species and how to differentiate them. I can also follow the taxonomic tree upward and browse it to see how many similar looking species there are that way.
8 Likes