The Impact of VAE Design on Latent Pose Representations for Diffusion-based Sign Language Production
For researchers in sign language production, this work highlights the importance of latent space properties beyond reconstruction metrics, but the findings are incremental as they confirm known issues in latent diffusion pipelines.
The paper investigates how VAE design choices for encoding sign pose sequences affect latent space structure and downstream diffusion-based sign language production, finding that generative performance (measured by BLEU) is sometimes better explained by latent space properties than by reconstruction accuracy alone.
Latent diffusion approaches to sign language production (SLP) rely on an initial stage that learns an encoding of sign pose sequences, enabling generative modeling in the resulting latent space. The autoencoder used in this stage is typically evaluated in terms of reconstruction quality using geometric metrics common in SLP. While informative, these metrics do not fully capture latent space properties that may influence the training and performance of the downstream generative model. In this work, we investigate how architectural and training objective design choices in a variational autoencoder (VAE) for sign pose encoding affect latent space structure, and how these differences translate into the performance of a latent diffusion model for text-to-sign generation. Our experiments on Phoenix14T dataset show that variations in generative performance, measured through back-translation BLEU scores, can sometimes be better explained by differences in latent space properties than by VAE reconstruction accuracy alone.