READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (12) · 9766-9774
DOI: 10.1609/aaai.v40i12.37940 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Haotian Wang und 9 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
The introduction of diffusion models has brought significant advances to the field of audio-driven talking head generation. However, the extremely slow inference speed severely limits the practical implementation of diffusion-based talking head generation models. In this study, we propose READ, a real-time diffusion-transformer-based talking head generation framework. Our approach first learns a spatiotemporal highly compressed video latent space via a temporal VAE, significantly reducing the token count to accelerate generation. To achieve better audio-visual alignment within this compressed latent space, a pre-trained Speech Autoencoder (SpeechAE) is proposed to generate temporally compressed speech latent codes corresponding to the video latent space. These latent representations are then modeled by a carefully designed Audio-to-Video Diffusion Transformer (A2V-DiT) backbone for efficient talking head synthesis. Furthermore, to ensure temporal consistency and accelerated inference in extended generation, we propose a novel asynchronous noise scheduler (ANS) for both the training and inference processes of our framework. The ANS leverages asynchronous add-noise and asynchronous motion-guided generation in the latent space, ensuring consistency in generated video clips. Experimental results demonstrate that READ outperforms state-of-the-art methods by generating competitive talking head videos with significantly reduced runtime, achieving an optimal balance between quality and speed while maintaining robust metric stability in long-time generation.
Zitieren
APA
Wang, H., Weng, Y., Du, J., Xu, H., Wu, X., He, S., Yin, B., Liu, C., Gao, J., & Liu, Q. (2026). READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation. Proceedings of the AAAI Conference on Artificial Intelligence, 40(12), 9766-9774. https://doi.org/10.1609/aaai.v40i12.37940
Vancouver
Wang H, Weng Y, Du J, Xu H, Wu X, He S, et al. READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(12):9766-9774. doi:10.1609/aaai.v40i12.37940
BibTeX
@article{Wang2026,
author = {Wang, Haotian and Weng, Yuzhe and Du, Jun and Xu, Haoran and Wu, Xiaoyan and He, Shan and Yin, Bing and Liu, Cong and Gao, Jianqing and Liu, Qingfeng},
title = {READ: Real-time and Efficient Asynchronous Diffusion for Audio-driven Talking Head Generation},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {12},
pages = {9766--9774},
doi = {10.1609/aaai.v40i12.37940}
}
EUVIMED