UniDDT: Unifying Multimodal Understanding and Generation with Decoupled Diffusion Transformer 文章

ArXiv CS.CV2026-06-16NEWSen作者: Shuai Wang, Liang Li, Yang Chen, Ruopeng Gao, Yao Teng, Limin Wang
查看原文 →
关系图谱

UniDDT: Unifying Multimodal Understanding and Generation with Decoupled Diffusion Transformer · 相关技术

相关技术

Noisy ViTLLMUnified Multimodal modelsblock-diffusion decoderNoisy ViT encoder