Skip to content
Hugging Face Daily Papers·· 16 days agoAI score28

NitroFusion2: Introducing Adversarial Loss into the Text Stream of Multimodal Diffusion Transformers

NitroFusion2: Bringing Adversarial Loss to the Text Stream of Multimodal Diffusion Transformers

AI briefing

NitroFusion2 reduces few-step generation of adversarial diffusion distillation from the usual 4-8 steps to 1-2 steps with comparable quality. It leverages the MMDiT text stream as a manipulable adversarial representation via Text-Stream Adversarial Distillation (TSAD) and Aspect-Cue Discrimination (ACD). One-step generation produces clearly rendered text and rich details, while two steps remain competitive on evaluation benchmarks.

Open original arxiv.org