GenSync: A Generalized Talking Head Framework for Audio-driven Multi-Subject Lip-Sync using 3D Gaussian Splatting

We introduce GenSync, a novel framework for multi-identity lip-synced video synthesis using 3D Gaussian Splatting. Unlike most existing 3D methods that require training a new model for each identity , GenSync learns a unified network that synthesizes lip-synced videos for multiple speakers. By incorporating a Disentanglement Module, our approach separates identity-specific features from audio representations, enabling efficient multi-identity video synthesis. This design reduces computational overhead and achieves 6.8x faster training compared to state-of-the-art models, while maintaining high lip-sync accuracy and visual quality.

@article{agarwal2025_2505.01928,
  title={ GenSync: A Generalized Talking Head Framework for Audio-driven Multi-Subject Lip-Sync using 3D Gaussian Splatting },
  author={ Anushka Agarwal and Muhammad Yusuf Hassan and Talha Chafekar },
  journal={arXiv preprint arXiv:2505.01928},
  year={ 2025 }
}