SnapGen++：释放扩散变换器潜力，在边缘设备上实现高效高保真图像生成

摘要

扩散变换器（DiT）的最新进展虽为图像生成设立了新标准，但由于其高昂的计算与内存成本，仍难以在终端设备上实际部署。本研究提出一种面向移动及边缘设备的高效DiT框架，可在严格资源限制下实现变换器级别的生成质量。我们的设计融合三大核心要素：首先，提出具有自适应全局-局部稀疏注意力机制的紧凑型DiT架构，平衡全局上下文建模与局部细节保留；其次，构建弹性训练框架，在统一超网络内联合优化不同容量的子DiT模型，使单一模型能动态适配多样化硬件的高效推理需求；最后，开发知识引导分布匹配蒸馏技术，该分步蒸馏流程将DMD目标与少步数教师模型的知识迁移相结合，生成适用于实时终端应用的高保真低延迟结果（如4步生成）。这些创新共同构建了可扩展、高效率、高质量的扩散模型，为多样化硬件部署提供支持。

English

Recent advances in diffusion transformers (DiTs) have set new standards in image generation, yet remain impractical for on-device deployment due to their high computational and memory costs. In this work, we present an efficient DiT framework tailored for mobile and edge devices that achieves transformer-level generation quality under strict resource constraints. Our design combines three key components. First, we propose a compact DiT architecture with an adaptive global-local sparse attention mechanism that balances global context modeling and local detail preservation. Second, we propose an elastic training framework that jointly optimizes sub-DiTs of varying capacities within a unified supernetwork, allowing a single model to dynamically adjust for efficient inference across different hardware. Finally, we develop Knowledge-Guided Distribution Matching Distillation, a step-distillation pipeline that integrates the DMD objective with knowledge transfer from few-step teacher models, producing high-fidelity and low-latency generation (e.g., 4-step) suitable for real-time on-device use. Together, these contributions enable scalable, efficient, and high-quality diffusion models for deployment on diverse hardware.

SnapGen++：释放扩散变换器潜力，在边缘设备上实现高效高保真图像生成

SnapGen++: Unleashing Diffusion Transformers for Efficient High-Fidelity Image Generation on Edge Devices

摘要

Support