Image de couverture de livre

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

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Liang, Zhanhao, Yang, Tao, Wu, Jie, Feng, Chengjian, Zheng, Liang
Format:	Preprint
Publié:	2026
Sujets:	Computer Vision and Pattern Recognition
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization
par: Liang, Zhanhao, et autres
Publié: (2024)

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation
par: Gu, Yuchao, et autres
Publié: (2026)

Consistency Trajectory Matching for One-Step Generative Super-Resolution
par: You, Weiyi, et autres
Publié: (2025)

Learning Few-Step Diffusion Models by Trajectory Distribution Matching
par: Luo, Yihong, et autres
Publié: (2025)

RenderFlow: Single-Step Neural Rendering via Flow Matching
par: Zhang, Shenghao, et autres
Publié: (2026)

Know Your Step: Faster and Better Alignment for Flow Matching Models via Step-aware Advantages
par: Yue, Zhixiong, et autres
Publié: (2026)

X-SAM: From Segment Anything to Any Segmentation
par: Wang, Hao, et autres
Publié: (2025)

Diffusion Models Are Innate One-Step Generators
par: Zheng, Bowen, et autres
Publié: (2024)

Aligning Step-by-Step Instructional Diagrams to Video Demonstrations
par: Zhang, Jiahao, et autres
Publié: (2023)

Diffusion Adversarial Post-Training for One-Step Video Generation
par: Lin, Shanchuan, et autres
Publié: (2025)

BiFM: Bidirectional Flow Matching for Few-Step Image Editing and Generation
par: Dai, Yasong, et autres
Publié: (2026)

Trajectory Consistency for One-Step Generation on Euler Mean Flows
par: Li, Zhiqi, et autres
Publié: (2026)

SlimFlow: Training Smaller One-Step Diffusion Models with Rectified Flow
par: Zhu, Yuanzhi, et autres
Publié: (2024)

SnapFlow: One-Step Action Generation for Flow-Matching VLAs via Progressive Self-Distillation
par: Luan, Wuyang, et autres
Publié: (2026)

SubFlow: Sub-mode Conditioned Flow Matching for Diverse One-Step Generation
par: Lin, Yexiong, et autres
Publié: (2026)

Setting-Matched and Semantics-Scaled Benchmarking of One-Step Generative Models Against Multistep Diffusion and Flow Models
par: Ravishankar, Advaith, et autres
Publié: (2026)

FlowSteer: Guiding Few-Step Image Synthesis with Authentic Trajectories
par: Ke, Lei, et autres
Publié: (2025)

FlowIID: Single-Step Intrinsic Image Decomposition via Latent Flow Matching
par: Singla, Mithlesh, et autres
Publié: (2026)

High-Order Matching for One-Step Shortcut Diffusion Models
par: Chen, Bo, et autres
Publié: (2025)

Optimizing Few-Step Generation with Adaptive Matching Distillation
par: Bai, Lichen, et autres
Publié: (2026)

Flow-GRPO: Training Flow Matching Models via Online RL
par: Liu, Jie, et autres
Publié: (2025)

Self-Evaluation Unlocks Any-Step Text-to-Image Generation
par: Yu, Xin, et autres
Publié: (2025)

SoFlow: Solution Flow Models for One-Step Generative Modeling
par: Luo, Tianze, et autres
Publié: (2025)

One Small Step in Latent, One Giant Leap for Pixels: Fast Latent Upscale Adapter for Your Diffusion Models
par: Razin, Aleksandr, et autres
Publié: (2025)

Aligning Few-Step Diffusion Models with Dense Reward Difference Learning
par: Zhang, Ziyi, et autres
Publié: (2024)

Self-Corrected Flow Distillation for Consistent One-Step and Few-Step Text-to-Image Generation
par: Dao, Quan, et autres
Publié: (2024)

TReFT: Taming Rectified Flow Models For One-Step Image Translation
par: Li, Shengqian, et autres
Publié: (2025)

SimFlow: Simplified and End-to-End Training of Latent Normalizing Flows
par: Zhao, Qinyu, et autres
Publié: (2025)

MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation
par: Fu, Yuxiang, et autres
Publié: (2025)

DiSA: Diffusion Step Annealing in Autoregressive Image Generation
par: Zhao, Qinyu, et autres
Publié: (2025)

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

One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation
par: Li, Jianze, et autres
Publié: (2025)

ClipGStream: Clip-Stream Gaussian Splatting for Any Length and Any Motion Multi-View Dynamic Scene Reconstruction
par: Liang, Jie, et autres
Publié: (2026)

Flash Diffusion: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation
par: Chadebec, Clément, et autres
Publié: (2024)

SeedVR2: One-Step Video Restoration via Diffusion Adversarial Post-Training
par: Wang, Jianyi, et autres
Publié: (2025)

ChestX-Reasoner: Advancing Radiology Foundation Models with Reasoning through Step-by-Step Verification
par: Fan, Ziqing, et autres
Publié: (2025)

DeepSound-V1: Start to Think Step-by-Step in the Audio Generation from Videos
par: Liang, Yunming, et autres
Publié: (2025)

Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
par: Liu, Tao, et autres
Publié: (2026)

Precise: SDE-Consistent Stochastic Sampling for RL Post-Training of Flow-Matching Models
par: Zou, Jade, et autres
Publié: (2026)

DiffMM: Efficient Method for Accurate Noisy and Sparse Trajectory Map Matching via One Step Diffusion
par: Han, Chenxu, et autres
Publié: (2026)