Sampling from diffusion probabilistic models (DPMs) is often expensive for high-quality image generation and typically requires many steps with a large model. In this paper, we introduce sampling Trajectory Stitching T-Stitch, a simple yet efficient technique to improve the sampling efficiency with little or no generation degradation. Instead of solely using a large DPM for the entire sampling trajectory, T-Stitch first leverages a smaller DPM in the initial steps as a cheap drop-in replacement of the larger DPM and switches to the larger DPM at a later stage. Our key insight is that different diffusion models learn similar encodings under the same training data distribution and smaller models are capable of generating good global structures in the early steps. Extensive experiments demonstrate that T-Stitch is training-free, generally applicable for different architectures, and complements most existing fast sampling techniques with flexible speed and quality trade-offs. On DiT-XL, for example, 40% of the early timesteps can be safely replaced with a 10x faster DiT-S without performance drop on class-conditional ImageNet generation. We further show that our method can also be used as a drop-in technique to not only accelerate the popular pretrained stable diffusion (SD) models but also improve the prompt alignment of stylized SD models from the public model zoo. Code is released at https://github.com/NVlabs/T-Stitch
University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Washington University in St. Louis(圣路易斯华盛顿大学);Georgia Institute of Technology(佐治亚理工学院);Carnegie Mellon University(卡内基梅隆大学);Cornell University(康奈尔大学)
STITCH-OPE: Trajectory Stitching with Guided Diffusion for Off-Policy Evaluation
Department of Computer Science, University of Toronto(多伦多大学计算机科学系);University of Toronto Robotics Institute(多伦多大学机器人研究所);Toyota Research Institute(丰田研究院);Vector Institute(向量研究所)
ReStNet: A Reusable & Stitchable Network for Dynamic Adaptation on IoT Devices
College of Computer science and Technology, Zhejiang University of Technology(浙江工业大学计算机科学与技术学院);Institute of Cyberspace Security, College of Information Engineering, Zhejiang University of Technology(浙江工业大学信息工程学院网络空间安全研究院);Binjiang Institute of Artificial Intelligence, Zhejiang University of Technology(浙江工业大学滨江人工智能研究院);College of Information and Communication Engineering, Harbin Engineering University(哈尔滨工程大学信息与通信工程学院);College of Telecommunications and Information Engineering, Nanjing University of Posts and Telecommunications(南京邮电大学通信与信息工程学院)