Enhancing Diffusion-Based Quantitatively Controllable Image Generation via Matrix-Form EDM and Adaptive Vicinal Training

2026-02-02

AI Summary

This work addresses the limitations of existing Continuous Conditional Diffusion Models (CCDMs), which suffer from outdated diffusion frameworks that compromise sampling efficiency and struggle to balance generation quality with speed. To overcome these challenges, we propose iCCDM, an improved CCDM framework that reformulates the Elucidated Diffusion Model (EDM) in matrix form for the first time, integrates an adaptive neighborhood training strategy, and introduces continuous regression-based label conditioning. The resulting approach significantly enhances both the quality and efficiency of quantitatively controllable image generation. Extensive experiments demonstrate that iCCDM outperforms state-of-the-art text-to-image models—including Stable Diffusion 3, FLUX.1, and Qwen-Image—across four benchmark datasets, achieving superior generation fidelity at a lower sampling cost.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Continuous Conditional Diffusion Model (CCDM) is a diffusion-based framework designed to generate high-quality images conditioned on continuous regression labels. Although CCDM has demonstrated clear advantages over prior approaches across a range of datasets, it still exhibits notable limitations and has recently been surpassed by a GAN-based method, namely CcGAN-AVAR. These limitations mainly arise from its reliance on an outdated diffusion framework and its low sampling efficiency due to long sampling trajectories. To address these issues, we propose an improved CCDM framework, termed iCCDM, which incorporates the more advanced \textit{Elucidated Diffusion Model} (EDM) framework with substantial modifications to improve both generation quality and sampling efficiency. Specifically, iCCDM introduces a novel matrix-form EDM formulation together with an adaptive vicinal training strategy. Extensive experiments on four benchmark datasets, spanning image resolutions from $64\times64$ to $256\times256$, demonstrate that iCCDM consistently outperforms existing methods, including state-of-the-art large-scale text-to-image diffusion models (e.g., Stable Diffusion 3, FLUX.1, and Qwen-Image), achieving higher generation quality while significantly reducing sampling cost.

Problem

• diffusion model
• controllable image generation
• sampling efficiency
• continuous conditioning
• image quality

Innovation

• Matrix-form EDM
• Adaptive Vicinal Training
• Continuous Conditional Diffusion Model
• Sampling Efficiency
• Quantitatively Controllable Image Generation