Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Yue, Zhong, Zhizhou, Liu, Minhao, Chen, Zhaokang, Wu, Bin, Zeng, Yubin, Zhan, Chao, He, Yingjie, Huang, Junxin, Zhou, Wenjiang
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2410.10122
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916662543908864
author Zhang, Yue
Zhong, Zhizhou
Liu, Minhao
Chen, Zhaokang
Wu, Bin
Zeng, Yubin
Zhan, Chao
He, Yingjie
Huang, Junxin
Zhou, Wenjiang
author_facet Zhang, Yue
Zhong, Zhizhou
Liu, Minhao
Chen, Zhaokang
Wu, Bin
Zeng, Yubin
Zhan, Chao
He, Yingjie
Huang, Junxin
Zhou, Wenjiang
contents Real-time video dubbing that preserves identity consistency while achieving accurate lip synchronization remains a critical challenge. Existing approaches face a trilemma: diffusion-based methods achieve high visual fidelity but suffer from prohibitive computational costs, while GAN-based solutions sacrifice lip-sync accuracy or dental details for real-time performance. We present MuseTalk, a novel two-stage training framework that resolves this trade-off through latent space optimization and spatio-temporal data sampling strategy. Our key innovations include: (1) During the Facial Abstract Pretraining stage, we propose Informative Frame Sampling to temporally align reference-source pose pairs, eliminating redundant feature interference while preserving identity cues. (2) In the Lip-Sync Adversarial Finetuning stage, we employ Dynamic Margin Sampling to spatially select the most suitable lip-movement-promoting regions, balancing audio-visual synchronization and dental clarity. (3) MuseTalk establishes an effective audio-visual feature fusion framework in the latent space, delivering 30 FPS output at 256*256 resolution on an NVIDIA V100 GPU. Extensive experiments demonstrate that MuseTalk outperforms state-of-the-art methods in visual fidelity while achieving comparable lip-sync accuracy. %The codes and models will be made publicly available upon acceptance. The code is made available at \href{https://github.com/TMElyralab/MuseTalk}{https://github.com/TMElyralab/MuseTalk}
format Preprint
id arxiv_https___arxiv_org_abs_2410_10122
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MuseTalk: Real-Time High-Fidelity Video Dubbing via Spatio-Temporal Sampling
Zhang, Yue
Zhong, Zhizhou
Liu, Minhao
Chen, Zhaokang
Wu, Bin
Zeng, Yubin
Zhan, Chao
He, Yingjie
Huang, Junxin
Zhou, Wenjiang
Computer Vision and Pattern Recognition
Real-time video dubbing that preserves identity consistency while achieving accurate lip synchronization remains a critical challenge. Existing approaches face a trilemma: diffusion-based methods achieve high visual fidelity but suffer from prohibitive computational costs, while GAN-based solutions sacrifice lip-sync accuracy or dental details for real-time performance. We present MuseTalk, a novel two-stage training framework that resolves this trade-off through latent space optimization and spatio-temporal data sampling strategy. Our key innovations include: (1) During the Facial Abstract Pretraining stage, we propose Informative Frame Sampling to temporally align reference-source pose pairs, eliminating redundant feature interference while preserving identity cues. (2) In the Lip-Sync Adversarial Finetuning stage, we employ Dynamic Margin Sampling to spatially select the most suitable lip-movement-promoting regions, balancing audio-visual synchronization and dental clarity. (3) MuseTalk establishes an effective audio-visual feature fusion framework in the latent space, delivering 30 FPS output at 256*256 resolution on an NVIDIA V100 GPU. Extensive experiments demonstrate that MuseTalk outperforms state-of-the-art methods in visual fidelity while achieving comparable lip-sync accuracy. %The codes and models will be made publicly available upon acceptance. The code is made available at \href{https://github.com/TMElyralab/MuseTalk}{https://github.com/TMElyralab/MuseTalk}
title MuseTalk: Real-Time High-Fidelity Video Dubbing via Spatio-Temporal Sampling
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.10122