NitroFusion2: Bringing Adversarial Loss to the Text Stream of Multimodal Diffusion Transformers
cs.CV
Submitted: 2026-11-03
Updated: 2026-11-03
Terminology
Sources
- eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers
- SD3.5-Flash: Distribution-Guided Distillation of Generative Flows
- X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again
- Multistep Consistency Models
- ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
- Distribution Matching Distillation without Fake Score Network
- 1.x-Distill: Breaking the Diversity, Quality, and Efficiency Barrier in Distribution Matching Distillation
- Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers
- Geometric GAN
- SDXL-Lightning: Progressive Adversarial Diffusion Distillation
- Self-Adversarial One Step Generation via Condition Shifting
- High-Fidelity Two-Step Image Generation via Teacher-Aligned End-to-End Distillation
- Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
- Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference
- AdvDMD: Adversarial Reward Meets DMD For High-Quality Few-Step Generation
- Qwen-Image Technical Report
- Diff-Instruct with Diffused Reward: Towards Principled One-step Generator RL
- One-step Diffusion Models with $f$-Divergence Distribution Matching
- Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs