Timezone: »

Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models
Fan Bao · Chongxuan Li · Jiacheng Sun · Jun Zhu · Bo Zhang

Tue Jul 19 08:00 AM -- 08:05 AM (PDT) @ Room 310

Diffusion probabilistic models (DPMs) are a class of powerful deep generative models (DGMs). Despite their success, the iterative generation process over the full timesteps is much less efficient than other DGMs such as GANs. Thus, the generation performance on a subset of timesteps is crucial, which is greatly influenced by the covariance design in DPMs. In this work, we consider diagonal and full covariances to improve the expressive power of DPMs. We derive the optimal result for such covariances, and then correct it when the mean of DPMs is imperfect. Both the optimal and the corrected ones can be decomposed into terms of conditional expectations over functions of noise. Building upon it, we propose to estimate the optimal covariance and its correction given imperfect mean by learning these conditional expectations. Our method can be applied to DPMs with both discrete and continuous timesteps. We consider the diagonal covariance in our implementation for computational efficiency. For an efficient practical implementation, we adopt a parameter sharing scheme and a two-stage training process. Empirically, our method outperforms a wide variety of covariance design on likelihood results, and improves the sample quality especially on a small number of timesteps.

Author Information

Fan Bao (Tsinghua University)
Chongxuan Li (Tsinghua University)
Jiacheng Sun (Huawei Noah's Ark Lab)
Jun Zhu (Tsinghua University)
Bo Zhang (Tsinghua University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors