• Ever thought it'd be cool to have your art, writing, or challenge runs featured on PokéCommunity? Click here for info - we'd love to spotlight your work!
  • Welcome to PokéCommunity! Register now and join one of the best fan communities on the 'net to talk Pokémon and more! We are not affiliated with The Pokémon Company or Nintendo.

[Sword & Shield] Tool for identifying pokémon shadows in max raid battles

  • 9
    Posts
    5
    Years
    • Seen Dec 28, 2019
    Hi everyone,

    I created a website tool which allows you to upload a picture of the screen showing the pokémon shadow before a max raid battle and it tells you which pokémon that is.

    The tool is called PokeRSI - Pokémon Raid Shadow Identifier, and it is available at: https://www.poke-rsi.com/. The source code is also available on GitHub, for those interested: https://github.com/zentos1/poke-rsi-js.

    Below is a picture of the website with one example picture:

    [PokeCommunity.com] Tool for identifying pokémon shadows in max raid battles


    I made it just for fun, but I thought it could also be interesting for someone else, so I decided to share it here.

    PS: I asked on the questions forum before, and they told me it was OK to post about it here.
     
    Last edited:
    In simple terms, it's how similar the shadow in the picture is to the shadow of the predicted pokémon.

    If you want a more technical explanation:
    - We first overlap the shadow of the picture with the shadow of the prediction,
    - then we count how many pixels the two shadows have in common (this is done by computing the IoU - Intersection over Union between the two images).
    - The IoU value is the prediction confidence.
     
    Wow this is so cool! Even though I have 100% confidence in the anime shadows, the sprite shadows are a little bit more difficult, and have found myself stumped more than a fewtimes... haha
    Thanks for making this!!~
     
    In simple terms, it's how similar the shadow in the picture is to the shadow of the predicted pokémon.

    If you want a more technical explanation:
    - We first overlap the shadow of the picture with the shadow of the prediction,
    - then we count how many pixels the two shadows have in common (this is done by computing the IoU - Intersection over Union between the two images).
    - The IoU value is the prediction confidence.
    Technical was what I wanted, haha. Thanks for that, it makes sense to me. Good idea on that implementation!
     
    Wow this is so cool! Even though I have 100% confidence in the anime shadows, the sprite shadows are a little bit more difficult, and have found myself stumped more than a fewtimes... haha
    Thanks for making this!!~

    You're welcome...^^...I I had a lot of fun creating it~
     
    Back
    Top