jlamoreaux/image-in-image

Intelligently composite one generated image into another with natural lighting and perspective matching, perfect for creating mockups and visualizations.

Public
2 runs

Readme

AI Image Compositor

Intelligently place content into scenes with natural lighting and perspective. Perfect for mockups, advertising visualizations, and creative compositions.

Quick Start

  1. Describe your scene: e.g., “desert highway at sunset”
  2. Describe overlay content: e.g., “rock band concert poster”
  3. Choose composition type: billboard, poster_on_wall, photograph_in_hands, etc.
  4. Run the model - it handles the rest!

Main Parameters

  • composition_type: How to place the content (billboard, poster, screen, etc.)
  • scene_description: The main scene/background
  • overlay_content: What to place in the scene
  • overlay_image: (Optional) Use your own image instead of generating one
  • style: Visual style (realistic, cinematic, vintage, etc.)
  • integration_quality: basic, high, or ultra

Examples

Billboard on Highway

  • Scene: “desert highway with mountains”
  • Overlay: “summer music festival advertisement”
  • Type: billboard

Poster in Room

  • Scene: “cozy coffee shop interior”
  • Overlay: “vintage travel poster”
  • Type: poster_on_wall

Photo in Hands

  • Scene: “person sitting on park bench”
  • Overlay: “family portrait”
  • Type: photograph_in_hands

Tips

  • Be specific with descriptions for better results
  • Higher quality settings take longer but look more realistic
  • The model automatically matches lighting between elements
  • Use “custom” type for creative placements