SOURCE-LINKED INTELLIGENCE

Accelerating Video Diffusion via Training-Free Trajectory Routing

arXiv · AI, language, vision and robotics · article · Sep 24, 2026 · UTC

Video diffusion is computationally expensive, as it requires executing a large model across many denoising steps. Even with step-distillation, inference remains expensive because every distilled step still requires a costly model evaluation. We present TRACK: TRajectory-Aware Capacity routing via top-K selection, a heterogeneous denoising strategy that switches between compatible large and small models at selected steps, reducing the average cost per denoising evaluation. The switching steps are determined using a calibration process. TRACK first rolls out a reference trajectory with the large

Read original source ↗ Open in workspace

recordType

paper

region

Global