Diffusion Models with Double Guidance: Generate with aggregated datasets

2025-05-19

AI Summary

To address controllable generation challenges arising from attribute inconsistency and absent joint annotations in multi-source heterogeneous dataset fusion, this paper proposes a dual-guided diffusion generative framework that operates without requiring fully conditioned joint annotations. Methodologically, we design an implicit conditional aggregation mechanism coupled with a decoupled gradient modulation strategy to achieve cross-dataset attribute alignment and multi-condition cooperative control. Technically, we introduce a novel dual-conditional guidance architecture that rigorously enforces constraints on multiple attributes—even in the absence of jointly annotated samples. Evaluated on molecular and image generation tasks, our approach achieves significantly higher conditional alignment accuracy and superior controllability under missing conditions compared to state-of-the-art baselines. This work overcomes a key bottleneck in controllable generation for multi-source data fusion and establishes a new paradigm for high-precision, multi-condition modeling under low annotation cost.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Creating large-scale datasets for training high-performance generative models is often prohibitively expensive, especially when associated attributes or annotations must be provided. As a result, merging existing datasets has become a common strategy. However, the sets of attributes across datasets are often inconsistent, and their naive concatenation typically leads to block-wise missing conditions. This presents a significant challenge for conditional generative modeling when the multiple attributes are used jointly as conditions, thereby limiting the model's controllability and applicability. To address this issue, we propose a novel generative approach, Diffusion Model with Double Guidance, which enables precise conditional generation even when no training samples contain all conditions simultaneously. Our method maintains rigorous control over multiple conditions without requiring joint annotations. We demonstrate its effectiveness in molecular and image generation tasks, where it outperforms existing baselines both in alignment with target conditional distributions and in controllability under missing condition settings.

Problem

• Merging datasets with inconsistent attributes causes missing conditions
• Conditional generative models struggle with joint multi-attribute controllability
• Existing methods fail when no samples contain all conditions

Innovation

• Diffusion Model with Double Guidance technique
• Generates without full condition annotations
• Enhances controllability in missing conditions