FAQ

Why is finding dwarf galaxies important?
Dwarf galaxies are among the most dark matter-dominated systems known, which makes them excellent laboratories for studying the nature of dark matter. You can learn more about this on the Research tab. Also, since only about 65 of these galaxies are known, each one receives a unique name and is listed on its on wikipedia page. !

Can you go back after seeing a candidate? What if I made a mistake?
Going back to change answers is not supported because it can bias the overall results when people adjust their responses after seeing others’ comments on the talk forums. We encourage everyone to give their best guess, even if they feel uncertain.

That said, you can review your last 20 classifications in the ‘Recents’ tab and share any interesting candidates in the talk forums, even if you accidentally misclassified them. Additionally, each image is classified by multiple people, so if you think you made a mistake, don’t worry; it will likely be balanced out by others’ responses.

How do I know if I am doing this right?
Don’t worry too much about making an occasional mistake as more than one volunteer will classify each candidate. If you are unsure, consult the Tutorial, Field Guide, and the Talk boards for help with classifying!

How many candidates should I do for it to be useful?
Don’t worry about completing all the classifications, every single one is valuable to us. Of course, the more classifications we receive, the better, but even just one classification can make a difference.

Are all the images shown made from real astronomical images?
Not all of them! While most images are based on real telescope observations, some include simulated dwarf galaxies that we've added. These simulated examples help us calibrate the Zooniverse classifications and give volunteers extra practice in identifying potential galaxies.

That said, you can review your last 20 classifications in the ‘Recents’ tab and share any interesting candidates in the talk forums, even if you accidentally misclassified them. Additionally, each image is classified by multiple people, so if you think you made a mistake, don’t worry; it will likely be balanced out by others’ responses.

Why do some images look weird?

Our DELVE search covers a large portion of the sky, but in some regions, telescope coverage is limited or certain detectors may not be functioning properly. If you receive an image like the one above, inspect it as you normally would and still look for over-densities near the center of the stellar density plot. Refer to field guide if you need help.

Why can't we use machine learning?
We are actively looking at using machine learning in the future. Our current software already removes many of the junk candidates, but many still get through. A much later goal for this project is to eventually train a model to classify the candidates, but first we need your help to build a large data set of reliably labelled images that the model can learn from.