Chicago Wildlife Watch Talk

Repeated shot

  • lesizz by lesizz

    This exact shot has appeared to me numerous times. I can tell by the cars in the background. I'm wondering how many other photos are presented to the same person numerous times. Doesn't seem an efficient way to go.

    Posted

  • ForestPreserve by ForestPreserve moderator

    Sometimes the cameras get stuck like this. Last season we had "Arthur", 5,858 images of the same stupid plant. Once there was a coyote, and there were two or three squirrels. It was interesting (?) in that you could watch Arthur grow.

    The complete unabridged Arthur:

    http://talk.chicagowildlifewatch.org/?_ga=1.176981989.1368728814.1410457495#/collections/CCHS000043

    Sometimes Arthur showed odd behavior:

    http://talk.chicagowildlifewatch.org/?_ga=1.176981989.1368728814.1410457495#/collections/CCHS000054

    Posted

  • llehrerlpzoo.org by llehrerlpzoo.org scientist

    Unfortunately, this could be a case of a camera over triggering, not the same photo over and over. I am familiar with this site and these cars are parked along the street. Hopefully something exciting will walk by!

    Posted

  • escholzia by escholzia

    Easy for you to say! Try clicking through hundreds (thousands?) of these boring suckers....

    Posted

  • DZM by DZM admin in response to ForestPreserve's comment.

    5,858 images

    I can't believe you COUNTED them. :p

    Posted

  • escholzia by escholzia

    OK, here are some rough calculations for this camera - I'm not as obsessive as ForestPreserve, so rough numbers will have to do. The images span the entire range from approximately ACH000eby2 to ACH000egkz. The last digit includes 36 images (0 - 9, a - z), the second to last 1296 (36 * 36) , and the third from the last, 233,280 (5 * 363636).. We can ignore everything but the biggest number. So, this camera took about 230,000 pictures of this tree, mostly without animals. That is FAR more than Arthur, and for 4 people to classify each one will require more than 500 hours of classifier time. I think I can speak for all volunteers that we get involved in these projects from scientific interest, not because we need robotic tasks to keep us busy. We had a discussion before this data set was posted, that the researchers would screen out cameras with more than 2000 pictures. Clearly that didn't happen here.

    Posted

  • ForestPreserve by ForestPreserve moderator in response to DZM's comment.

    I just collected them, trusting the count on the collection level. I used some very crude scripting to speed things up a bit. Not going to do that for this camera, based on escholzia's research.

    Posted

  • DZM by DZM admin in response to escholzia's comment.

    Hey @escholzia ... that does sound problematic. We are definitely not trying to waste people's time here.

    If your math is correct, we're talking about 1/3 of the entire data set being that same tree...

    But I'm not sure that it is correct. You did 5 * 36 * 36 * 36 ... but I think you may have added an extra *36?

    If we expand and imagine we're doing everything that is ACH000eb** and everything that is ACH000eg** ...

    There are 6 of the third-to-last digit (b, c, d, e, f, g)
    times 36 of the second to last digit
    times 36 of the last digit

    6 * 36 * 36 = 7776 photos. (And that's rounding up both the b** and the g**.) Similar to Arthur.

    Which is problematic -- still about 15 hours of classifier time -- but not quite the horror show you mention.

    Am I right on this?

    Posted

  • escholzia by escholzia

    Yes, you're right! Thanks for catching the mistake. But add the other 3-4 cameras with a similar issue, and it's still a burden. This one, for example: ACH000eglz.

    Posted

  • Bonnie123 by Bonnie123

    I too have had this picture so many times and never seen any animals. Is it really necessary for us to keep counting the bloody cars????? Could we just not hit the No animals button so that the image can be purged out? Could someone maybe explain to us what the added value is for counting the cars and humans? I think I've only had one image with a car in it where there was a squirrel/animal and I do not recall having any with a human that had any animals except a dog, so I'm not clear as to what the purpose is for having to count them.

    And escholzia is right with ACH000eglz I can't tell you how much I am starting to dislike That Tree!

    Posted

  • escholzia by escholzia

    They updated the info on the classification page to: Only tag cars that are specifically at the site. For example, a truck on a path would be included, while road traffic would be excluded. But experienced classifiers aren't likely to catch that change.

    I saw a squirrel here once ;-|

    Posted

  • Bonnie123 by Bonnie123

    @escholzia: Thanks so much for letting me know. You are right, I've been classifying for awhile now and I do not recall anything stating we only were to mark the cars at the site. To say the Zooniverse folks need to take some communication courses is an understatement. Thanks again for letting me know.

    Posted

  • escholzia by escholzia

    It was brought up in an earlier thread, but the change wasn't announced. However, I'm not complaining too much. I came here from Seafloor Explorer, which was abandoned by its founders and we had no way to contact anyone. The Chicago Wildlife Watch folks are really involved by comparison!

    There's a spectrum of responsibility toward volunteers across Zooinverse. Some projects get it that if you make volunteers partners in the science and keep them engaged intellectually you get much more out of the project (Galaxy Zoo) , while others see us as free labor and aren't too concerned about wasting our time. I think this project started small as community engagement for Chicago residents and they haven't given a lot of thought to what the volunteers get out of it. I stick with this because I enjoy not knowing what will be in the next pic - unless it's the damn Tree of course, but could wish for a better experience (fewer Trees) and more engagement in the science.

    Posted

  • Bonnie123 by Bonnie123

    @escholzia: You are correct the project did start small, when I first found it here it was buried under the Laboratory section (not on the main menu) and there were like only 600 volunteers working on it. I was so glad when they were able to put it on the main menu since it was more visable and there was a big increase in the volunteers. Totally agree with your comments about the volunteers some projects get it and some do not.

    I do admire the folks for Chicago Watch, I think it's rather nice to have a city care enough about how their wildlife are doing to go to all this effort. They really should be commended for all their efforts.

    They just started a new project on Chimps you might like it as well, I reviewed it before it launched and provided comments, it has lots of animals in it including elephants, hogs, and of course chimps. I hope if you try it you will stick with this as well, but it might provide a little variety for you.

    Posted

  • escholzia by escholzia

    I've done a bit of Chimps. It's an amazing project. All those duikers, though...

    Posted

  • DZM by DZM admin

    Some projects get it that if you make volunteers partners in the
    science and keep them engaged intellectually you get much more out of
    the project (Galaxy Zoo) , while others see us as free labor and
    aren't too concerned about wasting our time.

    Part of my role is to try to fix this paradigm. It frustrates me, too. You'll notice, I hope, that Chimp & See has a far, far more engaged science team. We're working on building a set of best practices for user engagement for new science teams to follow. C&S will provide a good blueprint.

    The car thing just came up out of Talk. We only just realized that it was causing confusion. So I'm not really sure what sort of announcement we should make... doesn't really merit an email. Hopefully updating the field guide will help.

    The science team is planning some science posts to mark the 1 million classification threshold. Those should be coming soon. Stay tuned!

    Posted

  • mason_UWI by mason_UWI scientist

    Hopefully the post will be coming out today! It's been done for a week, but I'm at the mercy of the communications department! Aside from that, we've been quite busy preparing for the urban wildlife conference we are hosting at the Lincoln Park Zoo this month and are waiting to hear back about two manuscripts we've submitted regarding this camera trap project. It takes a while to hear back on stuff like that, so we generally wait to announce 'cool' findings until we get the work published.

    Thanks so much for helping with the project!

    -Mason

    Posted

  • mason_UWI by mason_UWI scientist

    And the blog post is up! Thanks for all of your help everyone!

    Posted

  • DZM by DZM admin

    Excellent! Thank you so much for sharing!!

    Posted