cs.CVNov 19, 2026

TrajCal: Trajectory-Calibrated Time Schedules for Few-Step Flow Sampling

Authors: Junchi Zhang, Zihan Gu, Jiangcan Jia, Lulu Zhao, Wenbo Su, Hua Zhang

Organizations: Fudan University · Institute of Information Engineering, Chinese Academy of Sciences · School of Cyber Security, University of Chinese Academy of Sciences · East China Normal University · 

Abstract

Flow matching enables high-quality image and video generation, but its iterative sampling process requires repeated evaluations of large neural networks. Existing few-step methods can distort the output and reduce its agreement with a many-step reference. We propose TrajCal, a training-free calibration framework for few-step sampling. Based on scatter search, a well-established metaheuristic for difficult optimization problems, TrajCal starts with diverse seed schedules, deforms and combines these candidates, and selects one fixed schedule by final output fidelity for reuse on new inputs. For both image and video generation, TrajCal achieves higher reference fidelity than all compared baselines at low evaluation budgets while maintaining competitive visual quality. The search is effective even with randomly generated seeds, while trajectory-based seeds further improve fidelity. For cross-style generalization, photography-calibrated schedules also improve fidelity on unseen image styles.

Figures & tables

Appendix figures & tables20 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. ReAL: Accelerating Flow Matching through Segment Advancement with Shared Lookahead

    Sep 27, 2026Xuanhua Yin, Chuanzhi Xu, Haoxian Zhou +2Flow MatchingFlow-Based Generative Modeling

  2. CAT-Flow: Curvature-Adaptive sTeps for Flow Matching

    Sep 1, 2026Qinchan Li, Pedro Cisneros-Velarde, Keru Fu +3Flow MatchingT2I Generation

  3. Normalizing Trajectory Models

    May 8, 2026Jiatao Gu, Tianrong Chen, Ying Shen +3Normalizing FlowsDiffusion Model Distillation