Imagen de Portada

LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Liang, Zhanhao, Yang, Tao, Wu, Jie, Feng, Chengjian, Zheng, Liang
Formato:	Preprint
Publicado:	2026
Materias:	Computer Vision and Pattern Recognition
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization
por: Liang, Zhanhao, et al.
Publicado: (2024)

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation
por: Gu, Yuchao, et al.
Publicado: (2026)

Consistency Trajectory Matching for One-Step Generative Super-Resolution
por: You, Weiyi, et al.
Publicado: (2025)

Learning Few-Step Diffusion Models by Trajectory Distribution Matching
por: Luo, Yihong, et al.
Publicado: (2025)

RenderFlow: Single-Step Neural Rendering via Flow Matching
por: Zhang, Shenghao, et al.
Publicado: (2026)

Know Your Step: Faster and Better Alignment for Flow Matching Models via Step-aware Advantages
por: Yue, Zhixiong, et al.
Publicado: (2026)

X-SAM: From Segment Anything to Any Segmentation
por: Wang, Hao, et al.
Publicado: (2025)

Diffusion Models Are Innate One-Step Generators
por: Zheng, Bowen, et al.
Publicado: (2024)

Aligning Step-by-Step Instructional Diagrams to Video Demonstrations
por: Zhang, Jiahao, et al.
Publicado: (2023)

Diffusion Adversarial Post-Training for One-Step Video Generation
por: Lin, Shanchuan, et al.
Publicado: (2025)

BiFM: Bidirectional Flow Matching for Few-Step Image Editing and Generation
por: Dai, Yasong, et al.
Publicado: (2026)

Trajectory Consistency for One-Step Generation on Euler Mean Flows
por: Li, Zhiqi, et al.
Publicado: (2026)

SlimFlow: Training Smaller One-Step Diffusion Models with Rectified Flow
por: Zhu, Yuanzhi, et al.
Publicado: (2024)

SnapFlow: One-Step Action Generation for Flow-Matching VLAs via Progressive Self-Distillation
por: Luan, Wuyang, et al.
Publicado: (2026)

SubFlow: Sub-mode Conditioned Flow Matching for Diverse One-Step Generation
por: Lin, Yexiong, et al.
Publicado: (2026)

Setting-Matched and Semantics-Scaled Benchmarking of One-Step Generative Models Against Multistep Diffusion and Flow Models
por: Ravishankar, Advaith, et al.
Publicado: (2026)

FlowSteer: Guiding Few-Step Image Synthesis with Authentic Trajectories
por: Ke, Lei, et al.
Publicado: (2025)

FlowIID: Single-Step Intrinsic Image Decomposition via Latent Flow Matching
por: Singla, Mithlesh, et al.
Publicado: (2026)

High-Order Matching for One-Step Shortcut Diffusion Models
por: Chen, Bo, et al.
Publicado: (2025)

Optimizing Few-Step Generation with Adaptive Matching Distillation
por: Bai, Lichen, et al.
Publicado: (2026)

Flow-GRPO: Training Flow Matching Models via Online RL
por: Liu, Jie, et al.
Publicado: (2025)

Self-Evaluation Unlocks Any-Step Text-to-Image Generation
por: Yu, Xin, et al.
Publicado: (2025)

SoFlow: Solution Flow Models for One-Step Generative Modeling
por: Luo, Tianze, et al.
Publicado: (2025)

One Small Step in Latent, One Giant Leap for Pixels: Fast Latent Upscale Adapter for Your Diffusion Models
por: Razin, Aleksandr, et al.
Publicado: (2025)

Aligning Few-Step Diffusion Models with Dense Reward Difference Learning
por: Zhang, Ziyi, et al.
Publicado: (2024)

Self-Corrected Flow Distillation for Consistent One-Step and Few-Step Text-to-Image Generation
por: Dao, Quan, et al.
Publicado: (2024)

TReFT: Taming Rectified Flow Models For One-Step Image Translation
por: Li, Shengqian, et al.
Publicado: (2025)

SimFlow: Simplified and End-to-End Training of Latent Normalizing Flows
por: Zhao, Qinyu, et al.
Publicado: (2025)

MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation
por: Fu, Yuxiang, et al.
Publicado: (2025)

DiSA: Diffusion Step Annealing in Autoregressive Image Generation
por: Zhao, Qinyu, et al.
Publicado: (2025)

When Few Steps Are Enough: Training-Free Acceleration of Identity-Preserved Generation
por: Zheng, Dongqi
Publicado: (2026)

One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation
por: Li, Jianze, et al.
Publicado: (2025)

ClipGStream: Clip-Stream Gaussian Splatting for Any Length and Any Motion Multi-View Dynamic Scene Reconstruction
por: Liang, Jie, et al.
Publicado: (2026)

Flash Diffusion: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation
por: Chadebec, Clément, et al.
Publicado: (2024)

SeedVR2: One-Step Video Restoration via Diffusion Adversarial Post-Training
por: Wang, Jianyi, et al.
Publicado: (2025)

ChestX-Reasoner: Advancing Radiology Foundation Models with Reasoning through Step-by-Step Verification
por: Fan, Ziqing, et al.
Publicado: (2025)

DeepSound-V1: Start to Think Step-by-Step in the Audio Generation from Videos
por: Liang, Yunming, et al.
Publicado: (2025)

Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
por: Liu, Tao, et al.
Publicado: (2026)

Precise: SDE-Consistent Stochastic Sampling for RL Post-Training of Flow-Matching Models
por: Zou, Jade, et al.
Publicado: (2026)

DiffMM: Efficient Method for Accurate Noisy and Sparse Trajectory Map Matching via One Step Diffusion
por: Han, Chenxu, et al.
Publicado: (2026)