Similar to the IdentiFriday project I’d like to start a project for removing unknowns. For the first ever one how does 4:30 PM EST sound for everyone?
I’ll continue to work out the kinds, but how about Monday, Wednesday, Thursday, and Sunday weekdays at 4:30 and Sundays all day to try and tackle the unknowns? 14,000 Pages worth of identifications to be made might seem like a lot, but I’ve done 10 pages in one day before so I think as a group we can get it done with determination and constant vigilance!
When identifying “Unknowns” always try to add the best ID you are comfortable with without using the CV. Granted, I’m sure the CV could in fact help you get a better identification as it is constantly being improved, but to use it one must click on the observation. Honestly, I don’t know why that is the way it is, but it is and the way to most easily and efficiently go through unknowns would be to not have to click on every individual observation, but instead just do what you can do without the CV and not having to actually click on the observation. Next, when an observation is entirely made up of something a human made (e.g. a shops icon that includes a fish, etc.) you can confidently mark it with the identification Homo Sapiens.Warning: if you are unsure whether the observation was meant specifically for the human made object don’t mark it as human,instead simply ask the user what it is meant for and mark it as “reviewed” to get it out of your pool of unknowns.
@lynnharper@spiphany@lotteryd@dianastuder@kzoebel You all are just a few people I’ve seen show interest into identifying unknowns. If you’d be willing to help out here is the post so you don’t miss it!
…. Ooooh… I don’t think I can agree with such an approach.
For one thing, the Computer Vision continues to improve and is a valuable tool.
Just getting things out of the Unknown pool for its own sake is not adding that much value or knowledge. I have often IDd observations at a very high level, but only after considering my resources and knowledge (admittedly, not all that much).
That blanket rule about Human is really unfounded. If it, * the subject of the photo, appears to be not wild, that should be shown with the DQA. There’s no call to mark it human just because of an artifact in the picture. * If the subject of the photo is clearly a curb or book, by all means mark it Human.
Going thru Unknowns is a good opportunity to give feedback and educate the new users about better practices for using iNaturalist.
I have a half dozen or so auto-text entries that I can add for newcomers with a couple keystrokes that give some insight in to how to get more out of iNaturalist.
But there are lots of unknowns such as cultivated plants and urban weeds where photos also include prominent artifacts. Often it’s hard to be sure what subject the observer intended. I’d want to be sure that the artifact is the primary subject before adding an ID of Homo sapiens and I don’t think that’s what the OP was proposing.
Gotta be careful with getting into technicalities though. Technically, the presence of a photo uploaded to iNaturalist is a sign of Homo sapiens being present at every observation on this site
Please - if you are going to sweep Unknown into broad IDs, check carefully first that you don’t fall in the placeholder booby trap.
Yesterday for example was a missing species to flag for curation. If I had swept thru with - it’s a fern - next. That is where it would have stayed (I always hope that the observer has their placeholder text recorded where iNat cannot vanish it - but perhaps they are in the field without internet ??)
If tackling Unknowns it is good to start with a location where you recognise some obs. And if you clear ‘yours’ then the next wider location. For me Cape Peninsula (easy with MANY eyes) then Western Cape (few eyes) then Africa beyond my South Africa (that is harder, but see my rewarding example above)
Along the iNat journey you weigh up where you can add value, what is worth your time and effort. Different strokes for different folks.
Hey, thanks for the invite, but spring is really springing here in Massachusetts (so many skunk cabbage photos today!) and I’m still trying to wade through elderberry observations from eastern Canada from the past two or three years, so… Pray for rain. Or a nice 18-inch March blizzard. That’ll push spring back a few weeks and give me time to work on Unknowns. But good luck!
But, I ~think~ @cs16-levi-Levi was saying not to spend too much time looking at a picture to determine the subject. I got the sense they meant to avoid ~spending time~ thinking about an ambiguous photo and just to default to Human if there was a human-made artifact in the photo.
That would remove it from the needs ID pile, but it doesn’t seem like a fair treatment, if the subject of the photo ~wasn’t~ a human-made artifact, such as a curb, book, or stop sign, etc.
I was sort of just joking without thinking much about OP’s intention. If there’s an organism in the observation along with the human artifact, one should certainly focus on the organism not the artifact.
Right – there could be a small lizard or fly on one of those artifacts, which was the observer’s intended subject even though it does not appear very prominent.
Yesterday I sorted by date observed, ascending.
Cleared the oldest 10 pages of Unknowns - which is 90% bacteria and virus - Mark as Reviewed - not going to move. Did retrieve some placeholder text options.
Then the 10 oldest pages of plants. Now my notifications will roll in from Fox Glacier NZ and Nepal and Chile. Virtual world tour.
Back to my African heart today.
Hello,
How do you deal with all these landscape photos without any focus on any particular plant. Just mark them as good as can be (moves them to hobby quality) or coment to ask for the uploaders intention and then mark them as viewed?
Also what about observation where there are multike photos with different species and the uploader never split the observation.
@cs16-levi Could you please address the comments responding to your idea of:
“When going through unknowns if you don’t know what it could be just add the broadest topic. If it shows concrete, books, or anything non-animal created by human mark it as Human. CV suggestions take time and require clicking on the observation! For more speed and getting rid of the chance of a wrong id, identify it as best you can w/out CV and the chance of a wrong id.”
Protecting the data in the photos from being lost to the “casual” category is more important than just lowering the “Unknown” numbers. If an organism other than “human” is in a photo and is potentially identifiable by others but you don’t know what it is, then please just skip it rather than looking for a human artifact to focus on just for a numbers game. That may not be what you were advocating, but it would be nice if you would clarify.
Then the 10 oldest pages of plants. Now my notifications will roll in from Fox Glacier NZ and Nepal and Chile. Virtual world tour.
I always enjoy the “virtual world tour” aspect of going through unknowns. Sometimes I get to see something familiar that I didn’t realise could be found so far away, but most of the time I get to find out about some interesting species I’ve never seen before.
