A$^{2}$V-SLP: Alignment-Aware Variational Modeling for Disentangled Sign Language Production

TL;DR

This paper introduces A$^{2}$V-SLP, a novel alignment-aware variational framework for sign language production that learns disentangled, articulator-specific latent representations to improve motion realism and translation accuracy.

Contribution

It proposes a variational autoencoder with articulator-wise disentangled latent distributions and a gloss attention mechanism for enhanced sign language generation.

Findings

Achieves state-of-the-art back-translation performance

Improves motion realism in sign language synthesis

Outperforms deterministic latent regression methods

Abstract

Building upon recent structural disentanglement frameworks for sign language production, we propose A$^{2}$V-SLP, an alignment-aware variational framework that learns articulator-wise disentangled latent distributions rather than deterministic embeddings. A disentangled Variational Autoencoder (VAE) encodes ground-truth sign pose sequences and extracts articulator-specific mean and variance vectors, which are used as distributional supervision for training a non-autoregressive Transformer. Given text embeddings, the Transformer predicts both latent means and log-variances, while the VAE decoder reconstructs the final sign pose sequences through stochastic sampling at the decoding stage. This formulation maintains articulator-level representations by avoiding deterministic latent collapse through distributional latent modeling. In addition, we integrate a gloss attention mechanism to strengthen alignment between linguistic input and articulated motion. Experimental results show consistent gains over deterministic latent regression, achieving state-of-the-art back-translation performance and improved motion realism in a fully gloss-free setting.