Refine image-to-video prompts using ChatGPT and image context

Step-by-step

1. Copy the motion-focused prompt engineering template
2. Upload the starting image to ChatGPT
3. Paste the template into the chat
4. Input a short concept describing the desired action, such as an orbital shot of a female gladiator
5. Specify that the uploaded image is the starting point for the shot

Tips

• Upload the image to provide ChatGPT with additional context so it can see exactly what is in the scene
• Focus the prompt on how the character moves, how the environment changes, and how the camera moves rather than defining the character or environment

Highlights

“it's useful to add the image in because it helps chat GBT have additional context for working with our prompt design”