Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > comfyui-diversityboost

ComfyUI Extension: comfyui-diversityboost

Repo Name

ComfyUI-DiversityBoost

Author
facok (Account age: 1188 days)
Nodes
View all nodes(2)
Latest Updated
2026-04-25
Github Stars
0.03K

How to Install comfyui-diversityboost

Install this extension via the ComfyUI Manager by searching for comfyui-diversityboost
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter comfyui-diversityboost in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

comfyui-diversityboost Description

Comfyui-diversityboost enhances composition diversity in distilled diffusion models using a training-free frequency-domain phase injection technique.

ComfyUI-DiversityBoost Introduction

ComfyUI-DiversityBoost is an innovative extension designed to enhance the creative potential of AI-generated images by restoring composition diversity in distilled diffusion models. These models, while capable of producing high-quality images quickly, often suffer from a phenomenon known as "composition collapse." This means that different seeds tend to generate images with nearly identical layouts, limiting the variety and creativity of the outputs. ComfyUI-DiversityBoost addresses this issue by introducing a method that requires no additional training or model modifications, making it an accessible and efficient solution for AI artists seeking to diversify their image compositions.

How ComfyUI-DiversityBoost Works

At its core, ComfyUI-DiversityBoost employs a technique called polynomial frequency modulation combined with a Discrete Cosine Transform (DCT) composition push. Imagine your image as a musical composition where each pixel is a note. In a typical distilled model, the "music" tends to sound the same regardless of the initial "tune" or seed. ComfyUI-DiversityBoost acts like a conductor, subtly altering the frequency and arrangement of these notes to create a more varied and dynamic composition.

  1. Polynomial Frequency Modulation: This process smooths out the high-frequency components of the image, akin to adjusting the treble in a music track. It ensures that the essential structure of the image remains intact while allowing for more variation in the composition.
  2. DCT Composition Push: This step introduces a random low-frequency spatial field, which redistributes the "energy" across the image. Think of it as adding a new rhythm to the music, encouraging the model to explore different compositions with each seed. By applying these techniques, ComfyUI-DiversityBoost enables the model to reconstruct details in a way that varies with each seed, leading to a richer diversity of outputs.

ComfyUI-DiversityBoost Features

ComfyUI-DiversityBoost offers a range of features that allow you to customize the diversity and composition of your images:

  • Strength: Controls the amplitude of the composition push. A higher value results in more pronounced changes in composition.
  • Clamp: Sets the upper limit for the multiplicative scale factor, ensuring that the changes remain within a desired range.
  • Noise Type: Choose from pink, white, or blue noise to influence the frequency spectrum of the random DCT coefficients, affecting the overall texture and feel of the image.
  • DC Preserve: Allows you to maintain the original brightness of the image, balancing diversity with consistency.
  • Energy Compensate: Rescales the output to match the original image's root mean square (RMS) energy, preserving the overall intensity.
  • HF Factor: Adjusts the strength of high-frequency attenuation, controlling the level of detail in the image.
  • LF Factor: Amplifies low-frequency components, enhancing the depth and richness of the composition.
  • Transition: Modifies the shape of the polynomial transition, affecting how smoothly the changes are applied.
  • Schedule: Determines the timing of the frequency modulation and DCT push, with options for flat, linear, or cosine schedules.

ComfyUI-DiversityBoost Models

ComfyUI-DiversityBoost includes two nodes for backward compatibility:

  • Diversity Boost (V3): The recommended node, featuring polynomial frequency modulation and token-grid normalization. It is designed for optimal diversity with minimal side effects.
  • Diversity Boost (Legacy): Retains the original Butterworth Low Pass Filter (LPF) approach for those who prefer the previous method.

What's New with ComfyUI-DiversityBoost

The latest version, DiversityBoost V3, introduces significant improvements:

  • Polynomial Frequency Modulation: Provides a smoother and more continuous attenuation of high-frequency components.
  • DCT Composition Push: Enhances the diversity of compositions by redistributing energy across the image.
  • Improved Parameter Controls: Offers more precise customization options for AI artists to tailor the diversity to their liking. These updates make ComfyUI-DiversityBoost more effective and user-friendly, allowing for greater creative exploration.

Troubleshooting ComfyUI-DiversityBoost

If you encounter issues while using ComfyUI-DiversityBoost, here are some common solutions:

  • Images Look Too Similar: Increase the strength or adjust the noise type to introduce more variation.
  • Loss of Detail: Reduce the HF factor to retain more high-frequency details.
  • Brightness Issues: Use the DC preserve option to maintain the original brightness levels. For further assistance, consider exploring community forums or tutorials for additional tips and tricks.

Learn More about ComfyUI-DiversityBoost

To deepen your understanding of ComfyUI-DiversityBoost and its capabilities, explore the following resources:

  • Tutorials: Look for online tutorials that provide step-by-step guidance on using the extension effectively.
  • Community Forums: Join discussions with other AI artists to share experiences and solutions.
  • Documentation: Review detailed documentation for in-depth explanations of each feature and parameter. These resources will help you make the most of ComfyUI-DiversityBoost and unlock new creative possibilities in your AI art projects.

comfyui-diversityboost Related Nodes

RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

comfyui-diversityboost detailed guide | ComfyUI