提出ASBM框架，让扩散模型采样路径更直更高效
Efficient Generative Modeling beyond Memoryless Diffusion via Adjoint Schrödinger Bridge Matching
- 用能量先验引导数据迁移，构建最优耦合路径
- 采样步数减少一半仍保持高保真图像生成
- 适合追求高效生成的科研与工程应用
扩散模型常因无信息量的无记忆前向过程导致轨迹高度弯曲、得分目标噪声大。本文提出伴随薛定谔桥匹配（ASBM），通过两阶段方法恢复高维空间中的最优生成轨迹：首先将薛定谔桥前向动态视为耦合构造问题，基于数据到能量的采样视角，将数据迁移到由能量定义的先验分布；随后利用诱导出的最优耦合，以简单匹配损失监督后向生成动态。通过脱离无记忆假设，ASBM显著产生更直线、更高效的采样路径。相比已有方法，ASBM在高维数据上展现出明显提升的稳定性和效率。图像生成实验表明，ASBM在更少采样步数下仍能实现更高保真度。我们进一步通过蒸馏将最优轨迹转化为单步生成器，验证其有效性。
原文摘要 · Abstract (English)
Diffusion models often yield highly curved trajectories and noisy score targets due to an uninformative, memoryless forward process that induces independent data-noise coupling. We propose Adjoint Schrödinger Bridge Matching (ASBM), a generative modeling framework that recovers optimal trajectories in high dimensions via two stages. First, we view the Schrödinger Bridge (SB) forward dynamic as a coupling construction problem and learn it through a data-to-energy sampling perspective that transports data to an energy-defined prior. Then, we learn the backward generative dynamic with a simple matching loss supervised by the induced optimal coupling. By operating in a non-memoryless regime, ASBM produces significantly straighter and more efficient sampling paths. Compared to prior works, ASBM scales to high-dimensional data with notably improved stability and efficiency. Extensive experiments on image generation show that ASBM improves fidelity with fewer sampling steps. We further showcase the effectiveness of our optimal trajectory via distillation to a one-step generator.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。