How does EyeQ accurately correct skin tones?

After processing hundreds of millions of images and receiving valuable feedback from customers, EyeQ recognized the need for an accurate skin tone correction solution that reveals the beauty of all skin tones. We’re pleased to share that we’ve achieved this request through our latest Perfectly Clear AI technology, which includes AI skin tone correction. 

How Does AI Skin Tone Correction Work?

Our AI deep learning system has been trained on a full range of skin tones to ensure complete accuracy. Dark skin tones are notoriously captured incorrectly in cameras. To address this, we’ve trained our AI model to detect these skin tones and apply a targeted correction to ensure the most lifelike complexion possible.

You can see the results of AI Skin Tone correction through our demo software, Workbench. In the “edit” mode of Workbench, you’ll find the AI skin tone correction tool located in the right hand panel:

This tool functions as follows:

  1. When certain skin tones are detected, the skin tone tool checkbox is automatically checked.
  1. When detected, the AI skin tone slider is applied along with other corrections (see the right panel update) to ensure the most accurate skin tone appearance. 
  1. Corrections are applied based on which scene the image is categorized as. At the moment, our AI skin tone correction is applied for the following scenes: iAuto People; Professional Portraits; Portraits with Dark Background; Portraits with Light Background; and Outdoor Portraits.
  1. As always, you can adjust the level and type of corrections to achieve your desired outcome. You can also create custom presets based on your tweaked corrections. 

Note: If AI skin tone is not detected, it cannot be forced on. But when detected, you can toggle on/off to see the effect.

Keep Reading

QuickDesk Workflow Automation

QuickDesk can be integrated into various LAB workflow tools like ROES using the ‘manifest file’ system described below.  ROES Server can be configured...

Read more