The matrix to describe, memorize, analyze, investigate

How to describe a face with a prompt for Midjourney and DALL-E?

Today, I'll take the example of a face you'd like to describe precisely to Midjourney and DALL-E. If you do it right, the result will be exactly what you want.

Artificial intelligence is a tool that needs precise details to give you a maximum satisfaction.

Prepare your reflection

Caucasian reflexive looking at laptop screen, reflexing on work, businesswoman

There are many aspects that must be considered when writing a prompt to describe a face, such as features, expressions, emotions, dimensions, proportions, etc.

How to detail a face and structure the description for Midjourney and DALL-E?

beautiful woman human robot artificial intelligence

"The more precise you are, the more you'll be bluffed!"

The Observable.fr matrix lets you structure your prompts perfectly, optimize them and preserve them over time. You can also create an observation per project, or a matrix per project. The choice is yours - the only limit is your imagination.

Here is a example that you can easily add in the matrix. Thanks to Observable.fr, you can copy and paste the prompt directly into the AI's user interface. It's simple and very effective.
You can also modify your prompts at any time, adding or deleting information, to meet the changing needs of your project.

Example of a human face description

Structuring your prompt to describe a face to Midjourney and DALL-E.

  • Describing a face
    • Facial features
      • Awareness of features
        • Trait identification
          • Front
            • Width
            • Height
            • Form
          • Nose
            • Length
            • Form
            • Size
          • Mouth
            • Size
            • Form
            • Color
          • Cheeks
            • Roundness
            • Size
          • Menton
            • Point
            • Form
    • Features
      • Awareness of characteristics
        • Identification of characteristics
          • Eyes
            • Eye color
              • Brown
              • Blue
              • Green
              • Grey
            • Eye shape
              • Oval
              • Rounds
              • In almond
            • Eye size
              • Big
              • Medium
              • Small
            • Eyebrows
              • Color
              • Form
              • Thickness
          • Hair
            • Hair color
              • Black
              • Brown
              • Red
              • Blond
              • Grey
            • Hair length
              • Short
              • Medium
              • Long
            • Hairstyle
              • Smooth
              • Curly
              • Frisée
          • Ears
            • Form
              • Round
              • Pointed
              • Plate
            • Size
          • Pilosity
            • Beard
            • Moustache
            • Eyebrows
    • Accessories
      • Awareness of accessories
        • Identification of accessories
          • Glasses
            • Form
              • Round
              • Square
              • Rectangular
            • Size
            • Color
          • Jewelry
            • Earrings
            • Necklace

This gives the AI a very clear basis for understanding which features need to be identified to describe a face.

Adhesive note with Good Job text on a cork bulletin board

Oh, I'll stop there, but it's a perfect start for you to become a whiz at sketching, well, so to speak...

If you think of anything else, share your idea with everyone in comments, it's very valuable!

Leave a Reply

Observation model (available in 2023)

A template is a predefined pattern of observations created by another user that you can immediately use to save time and discover different approaches. Any user can create a template and choose whether or not to share it with the public. You can add a model to your library, modify it and adapt it for new uses. You can also add a model directly to a current study or to a new study.

A template is composed of the following information:

  • The objective of the observational study

  • Possible additional explanations

  • All named analysis objectives (tabs)

  • All first-level descriptors in each analysis (N:0)

  • The category of the model according to the type of observation

  • The pseudo of the creator of the study

en_US