Kelp Talk

Hello, Floating Forests!

  • DZM by DZM admin

    Hello, Floating Forests Zooites! I wanted to stop by and introduce myself…

    My name is Darren--DZM here on Talk--and I've just joined the Zooniverse team. I'm working on a few projects, first and foremost an overhaul of the Talk system. In particular, I’m interested in working to build a unified Zooniverse Talk within which individual project forums are nested.

    I don’t have any sort of biological, botanical or other scientific background, but I’m hoping to learn from everyone here! I’ll be around these boards as often as possible, looking for any other ways in which I can help out, so I’m always interested in hearing from you--questions, concerns, ideas, anything.

    In particular, if you have any additional ideas for how Talk can be improved, I'd love to know!

    See you all around!

    DZM

    Posted

  • elizabeth_s by elizabeth_s

    😃 I see you are indeed making the rounds. This is a fun project hope you enjoy it.

    Posted

  • Artman40 by Artman40

    There is one problem about talk page. If I add something to the text, I have to close the window and re-open it before I can edit it again. This doesn't apply to later edits though. I think search feature needs more bugfixing and overhauling though.

    Posted

  • DZM by DZM admin

    Both good points, @artman40. Thank you!

    Posted

  • Artman40 by Artman40

    Another thing: would it be better that if you add a hashtag, it would show up as a link without a hashtag symbol to make it more pleasing to the eye?

    Posted

  • DZM by DZM admin

    That's also an idea, although I think having the # sign might be important so that someone knows that if they're clicking it, they're going to the hashtag compilation page, not to a different outside link? I think people are getting used to seeing that little # sign; what do other folks think?

    Posted

  • Artman40 by Artman40

    http://talk.snapshotserengeti.org/#/boards/BSG0000001/discussions/DSG000107e?page=1&comment_id=5438045dc67bee15770006de Here is some ideas about what to do with identified mistaggings.

    Posted

  • Artman40 by Artman40

    One more thing: It would be nice if Google Maps link of Floating Forests talk page would show more zoomed-in map that it currently is.

    Posted

  • DZM by DZM admin

    If anyone would like to know a bit more about me and my role, there's a new Zooniverse blog post up from Chris Lintott introducing me.

    Feel free to check it out, and I'm happy to answer any additional questions that you might have! 😃

    Posted

  • Artman40 by Artman40

    One new suggestion: would it be possible to remove favourite from subject page in case I accidentally favourite it or put it in a wrong collection? It's a hassle for having to go to a collection page, press edit, find the right page, select it for deletion and delete it.

    Posted

  • HMB6EQUJ5 by HMB6EQUJ5

    Hey all! I'm a Zooite newbie helping out mainly on Galaxy and occasionally on Sea Floor...i just heard you getting close to 1,000,000 mark and I just read the intro/info and am ready to "dive" in to help....btw i'm north of Monterey, Ca USA along the coast. Peace

    Posted

  • DZM by DZM admin

    Hello @HMB6EQUJ5! Great to have you; welcome!

    Good luck with your "dive;" please let us know if you find anything cool, and how your experience is going!

    Posted

  • Artman40 by Artman40

    One thing: I also suggest you to see comments which announce kelp, especially if it's faint or very faint. These help you in cases where kelp in not that visible. Currently the images have bad contrast so kelp is usually much fainter than shown in tutorial and field guide.

    Posted

  • DZM by DZM admin

    Hi Artman, I'm not 100% sure I understand. Are you suggesting that new users look through images that have been classified as having kelp?

    Thank you!

    Posted

  • Artman40 by Artman40

    Yes and update field guide accordingly to show hard cases. This decreases the amount of false negatives greatly, therefore increasing accuracy.

    Posted

  • DZM by DZM admin

    Ah, you're saying that new users need to be shown more clear examples of kelp.

    Yes, I agree completely. In fact, we just did a makeover of the Worm Watch Lab project, and I pushed very hard for the guide to include more examples of egg-laying activity, so that classifications would be more accurate.

    If we ever did a makeover of Floating Forests, it would definitely be good to do something similar.

    Posted

  • Artman40 by Artman40

    But don't worry. Floating Forests is not the only project that needs that kind of improvement.

    One thing that would also help is to make the surrounding background colour in the project darker. This makes it much easier to spot very faint kelp.
    AKP000992a is one of the examples where kelp is just barely visible and is easier to spot when the background is dark.

    Posted

  • DZM by DZM admin

    Yes! This is actually something that we did on our Chicago Wildlife Watch project... on the Classify interface, there's a little crescent-moon icon in the top right that switches to "dark mode." (http://www.chicagowildlifewatch.org) We definitely want to do that on future projects as well.

    Posted

  • Artman40 by Artman40

    At least the dark ages are finally over for Floating Forests and I don't need to look the pictures very closely anymore. Late 2013/2014 pictures still look bleached. Is it because that data was collected from a different Landsat satellite?

    What are the estimated casualties in form of false negatives?

    Also, the end of Dark Age seems to have significantly boosted the amount of classifications per day due to more users joining in.

    Posted

  • DZM by DZM admin

    I wouldn't worry too much about false positives or negatives.

    At Zooniverse, we have found time and time and time again that the overall body of classifications nearly always points to the right answer (or, at least, the classification that a trained scientist would make), even when you'd think that bad classifications would skew the results.

    Posted

  • Artman40 by Artman40

    And about the most recently taken "bleached" images?

    Posted

  • DZM by DZM admin

    You'd have to ask one of the scientists... that's out of my knowledge base! 😃

    Posted

  • Artman40 by Artman40

    By the way, this dataset is now done. Waiting for new subjects to activate.

    Posted

  • KrisMead by KrisMead

    Hey guys and gals, I do a 30 mins Floating Forest Youtube show which I scroll trough the vast amount of sat images while playing background music. My YT Nick is SychosisPatient. Maybe in future might do a talk show over the images. (KrisMead)

    Posted