i want to get some images with different motifs in same style that i have a consulting…
- User
i want to get some images with different motifs in same style that i have a consulting image. how to do that? which AI should i choice?
- Assistant
Excellent question! This is one of the most powerful and fun use cases for AI image generation. The technique you're looking for is broadly called "style transfer" or "style reference," and several AI tools are excellent at it.
Here’s a breakdown of how it works, which AI tools to choose, and step-by-step instructions.
The Core Concept: How It Works
You're giving the AI two main inputs:
- Your Consulting Image (The Style Reference): The AI analyzes this image to understand its artistic style—colors, textures, lighting, line work, composition patterns, and overall mood.
- A Text Prompt (The New Motif): You describe the new subject or scene you want to create.
The AI then combines these, applying the style from your image to the new content from your text prompt.
Which AI Should You Choose? (Top Recommendations)
Here are the best options, ranked from easiest to most powerful.
1. Midjourney (The Best Choice for Simplicity & High Quality)
Midjourney is currently the king of easily replicating a style. Its "Style Reference" feature is incredibly effective and simple to use.
- Pros: Extremely high-quality results, very easy to use, excellent at capturing the nuance of a style.
- Cons: Requires a paid subscription, operates through the Discord app.
2. Stable Diffusion (The Most Powerful & Customizable Choice)
Stable Diffusion is an open-source model, which means you can run it on your own computer (if you have a good graphics card) or use various web services. It offers the most control.
- Pros: Free to run locally, ultimate control over every parameter, powerful extensions like IP-Adapter and ControlNet are specifically designed for this.
- Cons: Steep learning curve, can be complex to set up, requires powerful hardware for local use.
- User-Friendly Interfaces: Tools like Fooocus, Automatic1111, and ComfyUI make Stable Diffusion easier to use. Fooocus is especially great for beginners.
3. Leonardo.Ai (A Great Web-Based Alternative)
Leonardo.Ai is a fantastic middle ground. It's a web platform that offers many of the advanced controls of Stable Diffusion in a much more user-friendly package.
- Pros: Web-based and easy to start, offers a free daily allowance of credits, has a feature called "Image Guidance" which is perfect for style transfer.
- Cons: The best features and higher generation speeds require a paid plan.
Step-by-Step Guides for Each Platform
Let's say your consulting image is a watercolor painting of a cat and you want to create images of a "majestic lion" and a "bustling futuristic cityscape" in the same watercolor style.
How to Do It in Midjourney
This is the most straightforward method.
-
Upload Your Image: In your Discord chat with the Midjourney Bot, click the
+button and upload your consulting image. Press Enter to send it. -
Get the Image URL: Click on the uploaded image to open it, then right-click and select "Copy Image Link" (or "Copy Image Address").
-
Use the
--srefParameter: Now, type your prompt using the--sref(style reference) command.- Type
/imagine prompt: - Follow it with your new motif description.
- Add
--srefand paste the URL of your image.
Example Prompts:
/imagine prompt: a majestic lion roaring on a cliff --sref https://your-image-url.com/cat.png/imagine prompt: a bustling futuristic cityscape with flying vehicles --sref https://your-image-url.com/cat.png
- Type
-
Adjust Style Weight (Optional): You can control how strongly the style is applied using
--sw(style weight), from 0 to 1000. The default is 100.--sw 50(less style)--sw 750(more style)
Example:
/imagine prompt: a majestic lion --sref https://your-image-url.com/cat.png --sw 750
How to Do It in Stable Diffusion (using Fooocus - a simple UI)
Fooocus makes this very easy with its "Image Prompt" feature, which uses an advanced technique called IP-Adapter behind the scenes.
- Open Fooocus: Launch the application.
- Check the "Input Image" Box: At the bottom of the screen, make sure the "Input Image" checkbox is ticked. This will open up the image input options.
- Upload Your Image: Drag and drop your consulting image into the "Image Prompt" area.
- Set the Mode: Click the "Advanced" checkbox below the image. You will see options like "Image Prompt", "PyraCanny", "FaceSwap". Make sure "Image Prompt" is selected.
- Choose the "Style" Preset: Under the "Image Prompt" options, you will see a dropdown. Choose "Style" to tell the AI to focus on copying the artistic style.
- Write Your Prompt: In the main text prompt box at the bottom, write the description of your new motif.
- Prompt:
a majestic lion - Prompt:
a bustling futuristic cityscape
- Prompt:
- Generate: Click the "Generate" button. Fooocus will apply the style of your uploaded image to the new subject. You can adjust the "Stop At" and "Weight" settings for more control.
How to Do It in Leonardo.Ai
Leonardo's "Image Guidance" feature is perfect for this.
- Go to AI Image Generation: Log in to Leonardo.Ai and navigate to the image generation tool.
- Find Image Guidance: On the left-hand panel, find the "Image Guidance" section and toggle it on.
- Upload Your Image: Upload your consulting image here.
- Select a Mode: You will see several modes. For your goal, choose "Style Reference". This tells the AI to extract the aesthetic and apply it.
- Adjust Strength: Use the "Guidance Strength" (or "Stylization") slider to control how heavily the style is applied. A good starting point is around 0.5 - 0.7.
- Write Your Prompt: In the main prompt box at the top, describe your new motif.
- Prompt:
a majestic lion - Prompt:
a bustling futuristic cityscape
- Prompt:
- Generate: Click "Generate" and see the results.
Summary and Final Recommendation
- For Beginners or Quick, High-Quality Results: Midjourney is your best bet. The
--srefcommand is magical. - For Power Users Who Want Ultimate Control (and Free Software): Stable Diffusion (with an interface like Fooocus or Automatic1111) is the way to go.
- For a Balance of User-Friendliness and Control: Leonardo.Ai is an excellent and powerful web-based option.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.