AI Activity
Developing a Photo Story Using AI-Generated Images

Purpose: Generate AI images from story prompts and explore how context in prompts influences image generation by comparing different AI tools.

No. of. Classes

One class (1 hour) is enough to complete the activity. (Laptops/desktops: 10, Students strength: 15 to 20).

Materials Required

Laptop with Internet connection.

Exercises

Exercise: 1



  1. Select one story from the sample stories provided.
  2. Generate images for the story using:
  3. Compare the images generated by both tools and observe differences in consistency, characters and scene continuity.


Exercise: 2 (Optional)



  1. Create a Photo Story using the generated images and export the final video.


Solutions



  • Click here Link to view the solution for Exercise 1 – Perchance (without context)
  • Click here to view the solution for Exercise 1 – Gemini (with context)
  • Click here to view the solution for Exercise 2 – Photo Story
  • Teacher's Instruction:
    1. Prepare a set of short sample stories such as moral stories, children’s stories, fables, or simple nature-based stories.
    2. Demonstrate how the same story prompt can be given to different AI image generators and explain the idea of providing context in prompts.
    3. Show students how to use Perchance AI Image Generator without context and Gemini with detailed context before they begin the activity.
    4. Guide students to save all generated images in clearly named folders for easy comparison.
    5. Encourage students to closely observe differences in character appearance, background details, and consistency across images generated by different tools.
    6. Ensure students are familiar with the Photo Story app, including importing images, adding text, adjusting image borders, and exporting the video.
    7. Ask students whether characters remain consistent across multiple images when context is not provided.
    8. Discuss how providing context in prompts helps AI generate more consistent images in a story sequence.
    9. Does AI create images by imagination or by combining and remixing patterns it has learned from many images?
    10. In the same line as previous question, ask students - when a 'wild' or imaginary prompt is searched on the internet, why do we not get useful results but an AI image generator still produces an image? For instance, ask students to try wild or impossible prompts like dinosaur wearing astronaut suit floating on the Moon for generating images