Phasic Content Fusing Diffusion Model with Directional Distribution Consistency for Few-Shot Model Adaption
This addresses the challenge of few-shot model adaption for generative AI in data-scarce scenarios, representing an incremental improvement with novel components.
The paper tackles the problem of training generative models with extremely limited data (less than 10 samples), where existing methods overfit and degrade content, by proposing a phasic content fusing diffusion model with directional distribution consistency loss, resulting in improved capture of content, style, and local details and demonstrated superiority over state-of-the-art methods in few-shot generative model adaption tasks.
Training a generative model with limited number of samples is a challenging task. Current methods primarily rely on few-shot model adaption to train the network. However, in scenarios where data is extremely limited (less than 10), the generative network tends to overfit and suffers from content degradation. To address these problems, we propose a novel phasic content fusing few-shot diffusion model with directional distribution consistency loss, which targets different learning objectives at distinct training stages of the diffusion model. Specifically, we design a phasic training strategy with phasic content fusion to help our model learn content and style information when t is large, and learn local details of target domain when t is small, leading to an improvement in the capture of content, style and local details. Furthermore, we introduce a novel directional distribution consistency loss that ensures the consistency between the generated and source distributions more efficiently and stably than the prior methods, preventing our model from overfitting. Finally, we propose a cross-domain structure guidance strategy that enhances structure consistency during domain adaptation. Theoretical analysis, qualitative and quantitative experiments demonstrate the superiority of our approach in few-shot generative model adaption tasks compared to state-of-the-art methods. The source code is available at: https://github.com/sjtuplayer/few-shot-diffusion.