Timezone: »

 
Poster
FP-Diffusion: Improving Score-based Diffusion Models by Enforcing the Underlying Score Fokker-Planck Equation
Chieh-Hsin Lai · Yuhta Takida · Naoki Murata · Toshimitsu Uesaka · Yuki Mitsufuji · Stefano Ermon

Wed Jul 26 02:00 PM -- 03:30 PM (PDT) @ Exhibit Hall 1 #712
in Poster Session 3 »

Score-based generative models (SGMs) learn a family of noise-conditional score functions corresponding to the data density perturbed with increasingly large amounts of noise. These perturbed data densities are linked together by the Fokker-Planck equation (FPE), a partial differential equation (PDE) governing the spatial-temporal evolution of a density undergoing a diffusion process. In this work, we derive a corresponding equation called the score FPE that characterizes the noise-conditional scores of the perturbed data densities (i.e., their gradients). Surprisingly, despite the impressive empirical performance, we observe that scores learned through denoising score matching (DSM) fail to fulfill the underlying score FPE, which is an inherent self-consistency property of the ground truth score. We prove that satisfying the score FPE is desirable as it improves the likelihood and the degree of conservativity. Hence, we propose to regularize the DSM objective to enforce satisfaction of the score FPE, and we show the effectiveness of this approach across various datasets.

[ PDF]  [ Poster] 

Author Information

Chieh-Hsin Lai (Sony AI)
Yuhta Takida (Sony Group Corporation)
Naoki Murata (Sony Group Corporation)
Toshimitsu Uesaka (Sony Group Corporation)
Yuki Mitsufuji (Sony Group Corporation)
Stefano Ermon (Stanford University)

More from the Same Authors