Z-Magic: Zero-shot Multiple Attributes Guided Image Creator

2025-03-15

AI Summary

Existing personalized content generation methods suffer from contextual incoherence in multi-attribute customization and prohibitively high computational costs under zero-shot settings. Method: This paper proposes a zero-shot multi-attribute collaborative image generation framework. It reformulates the generation task from a multivariate conditional probability perspective—explicitly modeling inter-attribute conditional dependencies—and reveals their intrinsic connection to multi-task learning. The framework integrates conditional diffusion model adaptation, zero-shot transfer, and multi-task collaborative optimization to ensure semantic consistency across attributes. Contribution/Results: Experiments demonstrate that our method significantly outperforms state-of-the-art approaches under zero-shot conditions. It enables high-fidelity, semantically consistent image generation for arbitrary attribute combinations while substantially reducing computational overhead.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

The customization of multiple attributes has gained popularity with the rising demand for personalized content creation. Despite promising empirical results, the contextual coherence between different attributes has been largely overlooked. In this paper, we argue that subsequent attributes should follow the multivariable conditional distribution introduced by former attribute creation. In light of this, we reformulate multi-attribute creation from a conditional probability theory perspective and tackle the challenging zero-shot setting. By explicitly modeling the dependencies between attributes, we further enhance the coherence of generated images across diverse attribute combinations. Furthermore, we identify connections between multi-attribute customization and multi-task learning, effectively addressing the high computing cost encountered in multi-attribute synthesis. Extensive experiments demonstrate that Z-Magic outperforms existing models in zero-shot image generation, with broad implications for AI-driven design and creative applications.

Problem

• Enhances contextual coherence in multi-attribute image generation
• Addresses zero-shot setting for personalized content creation
• Reduces computing cost in multi-attribute synthesis

Innovation

• Zero-shot multi-attribute image generation
• Conditional probability for attribute coherence
• Multi-task learning reduces computing cost