The Altair Community is migrating to a new platform to provide a better experience for you. In preparation for the migration, the Altair Community is on read-only mode from October 28 - November 6, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here

Compare Faces with AWS Rekognition in RapidMiner Studio

sgenzersgenzer Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
edited November 2020 in Knowledge Base
This is a step-by-step guide for anyone to use RapidMiner Studio to leverage the capabilities of AWS Rekognition to compare two face images.

Part 1: Getting Started with AWS

1. Create an AWS Account. Full instructions are here. You will have to enter a payment method (credit card is preferred) but unless you do something really crazy, you will incur zero charges comparing a few faces using this method.

2. Sign into your new AWS account.

[2a. I strongly recommend enabling two-factor authentication at this point for your own protection. You can learn how to do this here.]

3. Go to Identity and Access Management (IAM) and create a new user for your RapidMiner tool.















*****This last step is VERY IMPORTANT!! KEEP THIS SCREEN OPEN. You will need these Access and Secret Keys in just a few minutes. *****

Part 2: Getting Started with RapidMiner

1. Download and install RapidMiner Studio. The full instructions are here.

2. Go to the top menu and select Extensions -> Marketplace



3. Search for "AWS Services". Click the checkbox for "Select for installation", accept the EULA, and restart RapidMiner Studio.

4. Go to the top menu and select Connections -> Create Connection



5. Choose "AWSServices", "Local Repository" and type "AWS Rekognition". Click "Create".



6. COPY and PASTE the Access and Secret keys from AWS into the boxes on this screen. Click "Save".



7. Drag and Drop this new "Connection" operator onto your Process panel.



8. Drag and Drop a "Compare Faces" operator onto your Process panel.



9. "Wire" up the two operators by clicking and dragging like this:



10. On the right side, select a file for your source image and another for your target image (it really doesn't matter which is which for most purposes).



11. Click the RUN button on the top left.



12. If all goes well, you will get a Similarity measure (from 0-100) as a result. That's it! :smile:



[NOTE: I strongly suggest at least doing a basic tutorial to get the basic idea of how the software works. There are Getting Started pages here.]

Comments

  • Telcontar120Telcontar120 RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
    @sgenzer this is a great walkthrough!  But I am not seeing the AWS Services extension available in the marketplace.  Is it restricted or still in beta?  Thanks.
    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
  • sgenzersgenzer Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
    hey @Telcontar120 yeah I tried to bury the KB but of course you saw it. 😀 It has not been released yet - hopefully in the next two weeks. Gotta go thru code review, etc..

    If you send me an email I can reply back with a beta .jar file...

    Scott

  • MartinLiebigMartinLiebig Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,533 RM Data Scientist
    @Telcontar120 it is online now :)


    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • rileyerileye Member Posts: 5 Contributor I
    edited December 2020
    Hi.  @sgenzer. I tried running this example and get an error: Error while loading partitions file from com/amazonaws/partitions/endpoints.json 

    https://docs.google.com/document/d/1gNT9qeX_7WRfwOIoTDQJjW_quyQOYrPToDspsSYjauU/edit?usp=sharing

    Anyone else got this, and a resolution?
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,533 RM Data Scientist
    Hi @rileye ,
    you are the second user reporting this to us. Unfortunately i was not able to reproduce the error on the images provided. May I ask what java version you are using and what your OS is?

    Best,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • rileyerileye Member Posts: 5 Contributor I
    Hi @mschmitz.  I'm using java version "1.8.0_271".
  • PhogenPhogen Member Posts: 2 Learner III
    Hi @rileye ,
    did you fix your error problem? If yes how?
    Thank you
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,533 RM Data Scientist
    Hi @Phogen ,
    there is a known issue in the current version. You can fix it by activating "Grant additional permissions to unsigned extensions" in your settings. We will release an update next week which fixes it.

    BR,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • PhogenPhogen Member Posts: 2 Learner III
    Thank you @mschmitz
    Now Compare face working OK. 
    I am a person who reported this bug in email to Scott Genzer  on December 9 and had email exchange with you and Scott.
    Thank you again!
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee-RapidMiner, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,533 RM Data Scientist
    HI @Phogen and @rileye ,

    I have just uploaded version 0.1.2 of the extension. This should fix the issues.

    I am sorry for the inconvenience.

    Best,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
Sign In or Register to comment.