LatentCRF: Continuous CRF for Efficient Latent Diffusion

TL;DR

LatentCRF is a neural network layer that models spatial and semantic relationships in latent diffusion models, significantly improving inference speed while maintaining image quality and diversity.

Contribution

We propose LatentCRF, a continuous CRF layer that replaces some inference steps in LDMs, enhancing efficiency without sacrificing quality.

Findings

33% faster inference with no quality loss

Maintains diversity and realism of generated images

Easy to integrate into existing LDMs

Abstract

Latent Diffusion Models (LDMs) produce high-quality, photo-realistic images, however, the latency incurred by multiple costly inference iterations can restrict their applicability. We introduce LatentCRF, a continuous Conditional Random Field (CRF) model, implemented as a neural network layer, that models the spatial and semantic relationships among the latent vectors in the LDM. By replacing some of the computationally-intensive LDM inference iterations with our lightweight LatentCRF, we achieve a superior balance between quality, speed and diversity. We increase inference efficiency by 33% with no loss in image quality or diversity compared to the full LDM. LatentCRF is an easy add-on, which does not require modifying the LDM.