Fast-DiM: Towards Fast Diffusion Morphs

Diffusion Morphs (DiM) are a recent state-of-the-art method for creating highquality face morphs; however, they require a high number of network functionevaluations (NFE) to create the morphs.We propose a new DiM pipeline, Fast-DiM,which can create morphs of a similar quality but with lower NFE. We investigatethe ODE solvers used to solve the Probability Flow ODE and the impact they haveon the the creation of face morphs. Additionally, we employ an alternativemethod for encoding images into the latent space of the Diffusion model bysolving the Probability Flow ODE as time runs forwards. Our experiments showthat we can reduce the NFE by upwards of 85experiencing only 1.6(MMPMR). Likewise, we showed we could cut NFE, in the sampling process, in halfwith only a maximal reduction of 0.23