So sánh phương pháp

	Mô hình Khuếch tán Đa phương thức	Mô hình khuếch tán tinh chỉnh
Lĩnh vực	Học sâu	Học sâu
Họ	Machine learning	Machine learning
Năm ra đời	2020–2022	2020–2023
Người khởi xướng	Ho, J. et al. (DDPM); Rombach, R. et al. (LDM/Stable Diffusion)	Ho, J., Jain, A., Abbeel, P. (base DDPM); Ruiz et al. (DreamBooth fine-tuning paradigm)
Loại	Generative model (denoising diffusion)	Generative model (fine-tuned via subject-specific or domain-specific data)
Công trình gốc	Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-Resolution Image Synthesis with Latent Diffusion Models. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10684–10695.	Ruiz, N., Li, Y., Jampani, V., Pritch, Y., Rubinstein, M., & Aberman, K. (2023). DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 22500–22510.
Tên gọi khác	multimodal DDPM, cross-modal diffusion, conditional multimodal diffusion, multi-modal denoising diffusion	DDPM fine-tuning, diffusion model adaptation, personalized diffusion model, subject-driven diffusion fine-tuning
Liên quan	6	5
Tóm tắt	A multimodal diffusion model extends denoising diffusion probabilistic models to generate or understand content by conditioning on signals from multiple modalities — such as text, image, audio, or video — simultaneously. It learns to reverse a noise process guided by cross-modal context, enabling high-fidelity synthesis and translation across modalities.	A fine-tuned diffusion model adapts a large pretrained denoising diffusion model — such as Stable Diffusion or DALL-E — to a specific subject, style, or domain by continuing training on a small curated dataset. Techniques such as DreamBooth, textual inversion, and LoRA make this adaptation feasible on consumer hardware while preserving general generative capability.