Multi-modal Latent Diffusion
This addresses the problem of generating coherent and high-quality multi-modal data for applications like multimedia analysis, though it is incremental as it builds on existing diffusion and autoencoder techniques.
The paper tackled the coherence-quality tradeoff in multi-modal Variational Autoencoders by proposing a method using independently trained uni-modal autoencoders and a masked diffusion model, achieving substantial improvements in both generation quality and coherence over competitors.
Multi-modal data-sets are ubiquitous in modern applications, and multi-modal Variational Autoencoders are a popular family of models that aim to learn a joint representation of the different modalities. However, existing approaches suffer from a coherence-quality tradeoff, where models with good generation quality lack generative coherence across modalities, and vice versa. We discuss the limitations underlying the unsatisfactory performance of existing methods, to motivate the need for a different approach. We propose a novel method that uses a set of independently trained, uni-modal, deterministic autoencoders. Individual latent variables are concatenated into a common latent space, which is fed to a masked diffusion model to enable generative modeling. We also introduce a new multi-time training method to learn the conditional score network for multi-modal diffusion. Our methodology substantially outperforms competitors in both generation quality and coherence, as shown through an extensive experimental campaign.