Sign in
Microsoft.com
 
Microsoft
 
 
 
Microsoft >
> 
> 
> 
Microsoft Face SDK for Cartoon Effects

Answered Microsoft Face SDK for Cartoon Effects

  • Friday, June 29, 2012 7:21 AM
     
     
    Sign In to Vote
    0

    Hi,

    Can we generate a cartoon using Face SDK. It should have some similarities to the original image which we passed has a parameter (  ImageConverter.SystemToSdk(image);  Here image is either from gallery or from the camera ) .  Now it is happening like , SDK has  some six different GUID vector categories based on the selected GUID category it is generating one random cartoon. It seems to looks like no relation with the original Image.

     

All Replies

  • Friday, June 29, 2012 8:41 AM
    Owner
     
     Answered
    Sign In to Vote
    0

    Several factors affect the cartoon effect, such as face pose, alignment result, illumination, the styles you choose, etc.

    • Pose

    Face pose will affect the final catoon figure. For example, if the input is a side-profile, then the output cartoon probably will also be a side-profile. A solution is to use a frontal, high-resolution face image.

    • Alignment result (face shape)

    Alignment result is a dominant factor for the cartoon face shape. Our alignemnt algorithm gives good result for most images, but still will fail on certain cases. It's recommended to manually calibrate each points of face shape to make it more accurate. This will improve the cartoon result.

    • Illumination

    In general, good lighting on the face part will result to good skin color on cartoon head.

    • Styles

    There six styles in cartoon. You can choose one that best matches with the input (e.g. skin color).

    In order to make a good cartoon result, I'd suggested involving human interfaction on certain steps.

     
  • Thursday, July 05, 2012 1:24 PM
     
     
    Sign In to Vote
    0

    Hi ,

    can you please suggest me how to find the color of the face using  face SDK

     
  • Tuesday, July 10, 2012 11:38 PM
    Owner
     
     
    Sign In to Vote
    0

    The released version does not have this interface. A workable approach is first generating a mask image based on the alignment result, and then calculating average face color over the mask region.