Curvature-Adaptive Consistency Flow Matching: Autonomous Trajectory Optimization via Reinforcement Learning
This work addresses the problem of inefficient training in consistency distillation for diffusion models, offering a dynamic curriculum that improves visual fidelity in few-step generation for large-scale models.
The paper identifies that consistency distillation faces optimization bottlenecks at boundary stages (initialization or final refinement) rather than intermediate steps, and proposes Curvature-Adaptive Consistency Flow Matching (CACFM) which uses a lightweight RL agent to dynamically construct an efficiency-oriented curriculum. The method achieves state-of-the-art results on FLUX and SDXL, mitigating structural deformities and preserving high-frequency details in few-step generation.
Consistency distillation has significantly accelerated the inference of diffusion models. In this work, we reveal an intriguing asymmetry: while Logit-Normal sampling priors are highly efficacious for standard iterative generation, consistency distillation exhibits a distinctly different difficulty profile (e.g., U-shaped). We identify that the primary optimization bottlenecks reside at the boundary stages (initialization or final refinement) rather than the intermediate steps. To address the limitations of static sampling in accommodating evolving learning requirements, we propose Curvature-Adaptive Consistency Flow Matching (CACFM). By formulating distillation as a dynamic decision process, CACFM employs a lightweight Reinforcement Learning agent to actively probe Probability Flow ODE trajectories, automatically constructing an efficiency-oriented curriculum that prioritizes critical regions without manual scheduling. Integrated with a novel Flow Distribution Matching Distillation (DMD) objective, our approach achieves new state-of-the-art results on large-scale models such as FLUX and SDXL. It effectively mitigates structural deformities and preserves high-frequency details in extreme few-step regimes, achieving unprecedented visual fidelity.