Orthogonal Adaptation for Modular Customization of Diffusion Models

• URL: http://arxiv.org/abs/2312.02432v3
• Date: Wed, 04 Dec 2024 22:39:52 GMT
• Title: Orthogonal Adaptation for Modular Customization of Diffusion Models
• Authors: Ryan Po, Guandao Yang, Kfir Aberman, Gordon Wetzstein,
• Abstract summary: We address a new problem called Modular Customization, with the goal of efficiently merging customized models. We introduce Orthogonal Adaptation, a method designed to encourage the customized models, which do not have access to each other during fine-tuning. Our proposed method is both simple and versatile, applicable to nearly all optimizable weights in the model architecture.
• Score: 39.62438974450659
• License:
• Abstract: Customization techniques for text-to-image models have paved the way for a wide range of previously unattainable applications, enabling the generation of specific concepts across diverse contexts and styles. While existing methods facilitate high-fidelity customization for individual concepts or a limited, pre-defined set of them, they fall short of achieving scalability, where a single model can seamlessly render countless concepts. In this paper, we address a new problem called Modular Customization, with the goal of efficiently merging customized models that were fine-tuned independently for individual concepts. This allows the merged model to jointly synthesize concepts in one image without compromising fidelity or incurring any additional computational costs. To address this problem, we introduce Orthogonal Adaptation, a method designed to encourage the customized models, which do not have access to each other during fine-tuning, to have orthogonal residual weights. This ensures that during inference time, the customized models can be summed with minimal interference. Our proposed method is both simple and versatile, applicable to nearly all optimizable weights in the model architecture. Through an extensive set of quantitative and qualitative evaluations, our method consistently outperforms relevant baselines in terms of efficiency and identity preservation, demonstrating a significant leap toward scalable customization of diffusion models.

Related papers

• Large Language Models to Diffusion Finetuning [20.251827725749607]
  We show our finetuned models achieve monotonically increasing accuracy, directly translating to improved performance across downstream tasks. Our method is universally applicable to any foundation model pre-trained with a cross-entropy loss.
  arXiv  Detail & Related papers  (2025-01-27T04:59:29Z)
• Merging Models on the Fly Without Retraining: A Sequential Approach to Scalable Continual Model Merging [75.93960998357812]
  Deep model merging represents an emerging research direction that combines multiple fine-tuned models to harness their capabilities across different tasks and domains. Current model merging techniques focus on merging all available models simultaneously, with weight matrices-based methods being the predominant approaches. We propose a training-free projection-based continual merging method that processes models sequentially.
  arXiv  Detail & Related papers  (2025-01-16T13:17:24Z)
• Training-free Heterogeneous Model Merging [40.681362819808136]
  We propose an innovative model merging framework designed for heterogeneous models. We show that the merging of structurally heterogeneous models can achieve performance levels comparable to those of homogeneous merging. Our code is publicly available at https://github.com/zju-vipa/training_free_heterogeneous_model_merging.
  arXiv  Detail & Related papers  (2024-12-29T04:49:11Z)
• LoRACLR: Contrastive Adaptation for Customization of Diffusion Models [62.70911549650579]
  LoRACLR is a novel approach for multi-concept image generation that merges multiple LoRA models, each fine-tuned for a distinct concept, into a single, unified model. LoRACLR uses a contrastive objective to align and merge the weight spaces of these models, ensuring compatibility while minimizing interference. Our results highlight the effectiveness of LoRACLR in accurately merging multiple concepts, advancing the capabilities of personalized image generation.
  arXiv  Detail & Related papers  (2024-12-12T18:59:55Z)
• Pareto Merging: Multi-Objective Optimization for Preference-Aware Model Merging [11.186194228460273]
  We propose a preference-aware model merging problem in which the performance of the merged model on each base model's task is treated as an objective. We show that the proposed model merging produces diverse trade-off models and achieves higher test accuracy compared to state-of-the-art merging baselines.
  arXiv  Detail & Related papers  (2024-08-22T03:41:14Z)
• JeDi: Joint-Image Diffusion Models for Finetuning-Free Personalized Text-to-Image Generation [49.997839600988875]
  Existing personalization methods rely on finetuning a text-to-image foundation model on a user's custom dataset. We propose Joint-Image Diffusion (jedi), an effective technique for learning a finetuning-free personalization model. Our model achieves state-of-the-art generation quality, both quantitatively and qualitatively, significantly outperforming both the prior finetuning-based and finetuning-free personalization baselines.
  arXiv  Detail & Related papers  (2024-07-08T17:59:02Z)
• Direct Consistency Optimization for Robust Customization of Text-to-Image Diffusion Models [67.68871360210208]
  Text-to-image (T2I) diffusion models, when fine-tuned on a few personal images, can generate visuals with a high degree of consistency. We propose a novel fine-tuning objective, dubbed Direct Consistency Optimization, which controls the deviation between fine-tuning and pretrained models. We show that our approach achieves better prompt fidelity and subject fidelity than those post-optimized for merging regular fine-tuned models.
  arXiv  Detail & Related papers  (2024-02-19T09:52:41Z)
• Multi-Concept Customization of Text-to-Image Diffusion [51.8642043743222]
  We propose Custom Diffusion, an efficient method for augmenting existing text-to-image models. We find that only optimizing a few parameters in the text-to-image conditioning mechanism is sufficiently powerful to represent new concepts. Our model generates variations of multiple new concepts and seamlessly composes them with existing concepts in novel settings.
  arXiv  Detail & Related papers  (2022-12-08T18:57:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.