The intrinsic link between facial motion and speech is often overlooked in generative modeling, where talking head synthesis and text-to-speech (TTS) are typically addressed as separate tasks. This paper introduces JAM-Flow, a unified framework to simultaneously synthesize and condition on both facial motion and speech. Our approach leverages flow matching and a novel Multi-Modal Diffusion Transformer (MM-DiT) architecture, integrating specialized Motion-DiT and Audio-DiT modules. These are coupled via selective joint attention layers and incorporate key architectural choices, such as temporally aligned positional embeddings and localized joint attention masking, to enable effective cross-modal interaction while preserving modality-specific strengths. Trained with an inpainting-style objective, JAM-Flow supports a wide array of conditioning inputs-including text, reference audio, and reference motion-facilitating tasks such as synchronized talking head generation from text, audio-driven animation, and much more, within a single, coherent model. JAM-Flow significantly advances multi-modal generative modeling by providing a practical solution for holistic audio-visual synthesis. project page: https://joonghyuk.com/jamflow-web
@article{arxiv.2506.23552,
title = {JAM-Flow: Joint Audio-Motion Synthesis with Flow Matching},
author = {Mingi Kwon and Joonghyuk Shin and Jaeseok Jung and Jaesik Park and Youngjung Uh},
journal= {arXiv preprint arXiv:2506.23552},
year = {2026}
}
Comments
project page: https://joonghyuk.com/jamflow-web Under review. Preprint published on arXiv