Improved Training Technique for Shortcut Models

Anh Nguyen, Viet Van Nguyen, Duc Vu, Trung Tuan Dao, Chi Tran, Toan Tran, Anh Tuan Tran

Abstract

Shortcut models represent a promising, non-adversarial paradigm for generative modeling, uniquely supporting one-step, few-step, and multi-step sampling from a single trained network. However, their widespread adoption has been stymied by critical performance bottlenecks.

This paper tackles the five core issues that held shortcut models back:

the hidden flaw of compounding guidance, which we are the first to formalize, causing severe image artifacts;
inflexible fixed guidance that restricts inference-time control;
a pervasive frequency bias driven by a reliance on low-level distances in the direct domain, which biases reconstructions toward low frequencies;
divergent self-consistency arising from a conflict with EMA training; and
curvy flow trajectories that impede convergence.

To address these challenges, we introduce iSM, a unified training framework that systematically resolves each limitation. Our framework is built on four key improvements: Intrinsic Guidance provides explicit, dynamic control over guidance strength, resolving both compounding guidance and inflexibility. A Multi-Level Wavelet Loss mitigates frequency bias to restore high-frequency details. Scaling Optimal Transport (sOT) reduces training variance and learns straighter, more stable generative paths. Finally, a Twin EMA strategy reconciles training stability with self-consistency.

Extensive experiments on ImageNet 256x256 demonstrate that our approach yields substantial FID improvements over baseline shortcut models across one-step, few-step, and multi-step generation, making shortcut models a viable and competitive class of generative models.