Back/Personal/Midjourney
BeginnerPersonal

How to Generate Realistic AI Self-Portraits for Content

Create recognizable self-portraits for articles and thumbnails by using a varied set of honest selfies as identity references, then combining that identity with a strong scene, pose, and expression.

How to Generate Realistic AI Self-Portraits for Content

Jamey loads several realistic selfies into Flora, uses them as identity references, and remixes strong Midjourney compositions with her face and the expression needed for an article.

Before you start

What you need

  • Six to twelve clear selfies you own
  • A reference-capable image tool such as Flora
  • A scene or composition brief
  • Optional Midjourney source image

What you’ll make

A set of realistic, recognizable portraits that fit specific content concepts while retaining a consistent identity.

Tools used

Step by step

The workflow

Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.

4 steps

Step01

Create a Diverse Selfie Reference Set

Collect a varied identity set with front, three quarter, and side angles, neutral and expressive faces, different lighting, and visible distinctive features. Use images you own and are comfortable processing.

Step02

Use Selfies as Image References

Upload the set as strong identity references in Flora or a similar tool. Run a neutral portrait first to confirm the model recognizes you before adding a complex scene.

Step03

Generate Yourself into New Scenarios

Describe the content concept, setting, pose, expression, camera treatment, and intended crop. Keep identity references separate from style or composition references so you can diagnose which input causes drift.

Example prompt
Create a realistic editorial portrait of me sitting at a desk, looking thoughtfully at a laptop. Use my uploaded selfies only for identity. Use the attached composition reference for pose and crop. Preserve my natural facial proportions and distinctive features. Style it like a contemporary Dazed editorial photo shoot.
Step04

Remix Existing Images with Your Face

For a stronger starting composition, import a Midjourney image and replace only the subject identity. Lock the pose, expression, lighting, background, and crop, then compare the result with several selfie references.

Example prompt
Replace the subject’s face with my identity from the uploaded selfie references. Keep the annoyed expression, head angle, pose, lighting, background, color treatment, and crop unchanged.

This technique is useful for creating custom reaction images or placing yourself into a pre-existing, well-composed scene.

What good looks like

  • Face shape and distinctive features remain recognizable across outputs.
  • The requested expression reads naturally.
  • The source composition and visual style survive a face remix.
  • The image is clearly generated and used with appropriate consent and disclosure.

Build your next product with ChatPRD

Turn an idea into a PRD, user stories, and a plan.

Try ChatPRD free

After the steps

Runbook notes

How to recover when the loop fails and where human judgment helps.

Recover

If it goes sideways

The generated face looks attractive but not recognizable
Add references with varied angles, neutral lighting, teeth, and distinctive features rather than only polished profile photos.
Identity transfer changes the composition or expression
Use the original image as the composition reference and name the exact expression, camera angle, and elements to preserve.
The model repeatedly changes age, skin tone, or facial structure
Strengthen diverse identity references, state the attributes to preserve, and reject outputs rather than reinforcing the drift.

Start shipping
better products.

Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.

Free to start
No credit card
SOC 2 certified
Enterprise ready