Learning Timeline
Key Insights
PhotoMaker's Advantage Over Dreambooth
Unlike Dreambooth, which requires at least 10 images for training, PhotoMaker can deliver excellent results using just a single image.
More Than Just a Face Swap
PhotoMaker doesn't just crudely cut and paste faces (face swap). It integrates your face directly into the AI model seamlessly, ensuring the lighting and texture look more realistic.
Image Quality Tip
While one image is enough, uploading multiple images from various angles will help the AI produce a result that more closely resembles your original features.
Prompts
Indiana Jones Example Prompt
Target:
PhotoMaker
Indiana Jones collecting a golden statue img
Circus Performer Example Prompt
Target:
PhotoMaker
An impressed performer in a circus juggling knives img
Step by Step
How to Generate Self-Portraits in Various Scenarios with PhotoMaker
- Visit the Replicate website and search for the 'PhotoMaker' model, or open it via the link provided in the description.
- Upload a photo of your face by clicking on the 'Input Image' section or take a 'live' photo using your camera.
- Ensure the uploaded photo has a clear facial expression (e.g., surprised or smiling).
- Navigate to the 'Prompt' section to enter the description of the image you want to generate.
- Write a description of your desired scenario (e.g., 'Indiana Jones collecting a golden statue').
- You must include the keyword 'img' or 'IMG' at the end of your prompt so the AI knows to use the face reference you uploaded.
- Select your desired 'Style' from the menu (e.g., 'Cinematic', 'Disney Character', or 'Photographic').
- Click the 'Submit' button to start the image generation process.
- Wait for the process to complete and see your face seamlessly blended into the scenario.
- If you aren't satisfied with the result, you can click 'Submit' again to regenerate a new variation.
