Optimizing for the Shortest Path in Denoising Diffusion Model
This work addresses efficiency and quality issues in diffusion models for applications like interactive data generation, representing an incremental improvement over existing methods.
The paper tackles the problem of slow denoising in diffusion models by proposing a shortest-path approach to optimize residual propagation, resulting in significantly reduced diffusion steps and improved visual fidelity in generated samples.
In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF.