The AI Influencer Creation Roadmap

The process of creating an AI influencer is broken down into three main steps:

Image Generation: The first step involves creating the initial AI influencer using a text prompt.
Consistency and Variation: You’ll then create different images of the same character by changing their poses, expressions, clothing, and location.
Video Conversion: The final step is to convert these images into realistic, engaging videos for social media.

Step 1: Generating the Initial AI Influencer

Using Fooocus AI

The video recommends using Fooocus AI, a free tool accessed through a Google Research Colab link. Once you open the link, connect to the GPU and click “Run Anyway.” This initiates a process that takes a few minutes to generate a unique link to the Fooocus AI interface.

Crafting the Prompt

The key to creating a great AI influencer is a well-crafted prompt. The video suggests using ChatGPT to generate a detailed prompt that describes the character’s appearance, style, and background.

Adjusting Settings

Before generating the image, you can optimize the settings in Fooocus AI for better results:

Advanced Box: Click this to access additional settings.
Quality: Select your desired quality.
Aspect Ratio: For Instagram, choose the 1:1 ratio.
Image Number: You can generate up to 32 images at once. The video recommends starting with four.
Negative Prompt: Use this box to add words you want to exclude from the generated image (e.g., “blurry,” “distorted”).

After setting up, paste your prompt and click “Generate.” Fooocus will create multiple images based on your specifications. Download the image you like best to use as your base character.

Step 2: Creating Consistent Variations

Maintaining a consistent look for your AI influencer is crucial for building a brand. The video demonstrates how to use the same base image to generate new pictures with different poses, expressions, and outfits.

Generating Variations

Disable Randomization: Uncheck the “Random” button to ensure the new images are based on your selected character.
Input Image: Go to the “Input” button, select “Image Prompt,” and upload your base character image.
Faceswap: In the “Advanced” settings, select “Faceswap” to ensure the new images retain the same face.
Modify the Prompt: Change elements in your prompt like clothing color, background, or location to create new scenes. For example, changing “pastel pink blazer” to “yellow kurta” and adding “a market scene” to the background.

Changing Poses and Expressions

Pose Implementation: To change the pose, upload a separate pose reference image in the right-side box. Make sure the “FaceSwap” and either the “Pyran” or “CPDS” options are selected in the advanced settings. Click “Generate” to create your new image.
Expression Variation: To change expressions, use the “Upscale or Variation” option. Upload your image, select “Very Strong,” and modify the prompt to describe the new expression, like changing “smiling” to “serious.”

Step 3: Converting Images to Videos

To increase engagement on platforms like Instagram, you can convert your images into short, dynamic videos. The video suggests using Google AI Studio, a free tool.

Using Google AI Studio

Access the Tool: Search for “Google AI Studio” and log in to the platform.
Upload Image: Navigate to “Generate Media” and upload your desired image.
Add a Simple Prompt: Enter a simple prompt describing the scene, like “A girl walking in a park.”
Generate Video: Click “Run” to start the video generation process. This can take a few minutes.

While Google AI Studio is a powerful tool, it has a limitation of generating only 3-4 videos per day. You can work around this by using multiple Google accounts.

Tools Used

Fooocus AI: A free AI tool for generating consistent images of a character.
Google AI Studio: A free tool for converting still images into realistic videos.
ChatGPT: An AI chatbot for creating detailed image prompts.

tutorao.com