Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zhu, Zeyu, Wu, Weijia, Shou, Mike Zheng
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2508.03050
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908478761598976
author Zhu, Zeyu
Wu, Weijia
Shou, Mike Zheng
author_facet Zhu, Zeyu
Wu, Weijia
Shou, Mike Zheng
contents Existing studies on talking video generation have predominantly focused on single-person monologues or isolated facial animations, limiting their applicability to realistic multi-human interactions. To bridge this gap, we introduce MIT, a large-scale dataset specifically designed for multi-human talking video generation. To this end, we develop an automatic pipeline that collects and annotates multi-person conversational videos. The resulting dataset comprises 12 hours of high-resolution footage, each featuring two to four speakers, with fine-grained annotations of body poses and speech interactions. It captures natural conversational dynamics in multi-speaker scenario, offering a rich resource for studying interactive visual behaviors. To demonstrate the potential of MIT, we furthur propose CovOG, a baseline model for this novel task. It integrates a Multi-Human Pose Encoder (MPE) to handle varying numbers of speakers by aggregating individual pose embeddings, and an Interactive Audio Driver (IAD) to modulate head dynamics based on speaker-specific audio features. Together, these components showcase the feasibility and challenges of generating realistic multi-human talking videos, establishing MIT as a valuable benchmark for future research. The code is avalibale at: https://github.com/showlab/Multi-human-Talking-Video-Dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2508_03050
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multi-human Interactive Talking Dataset
Zhu, Zeyu
Wu, Weijia
Shou, Mike Zheng
Computer Vision and Pattern Recognition
Existing studies on talking video generation have predominantly focused on single-person monologues or isolated facial animations, limiting their applicability to realistic multi-human interactions. To bridge this gap, we introduce MIT, a large-scale dataset specifically designed for multi-human talking video generation. To this end, we develop an automatic pipeline that collects and annotates multi-person conversational videos. The resulting dataset comprises 12 hours of high-resolution footage, each featuring two to four speakers, with fine-grained annotations of body poses and speech interactions. It captures natural conversational dynamics in multi-speaker scenario, offering a rich resource for studying interactive visual behaviors. To demonstrate the potential of MIT, we furthur propose CovOG, a baseline model for this novel task. It integrates a Multi-Human Pose Encoder (MPE) to handle varying numbers of speakers by aggregating individual pose embeddings, and an Interactive Audio Driver (IAD) to modulate head dynamics based on speaker-specific audio features. Together, these components showcase the feasibility and challenges of generating realistic multi-human talking videos, establishing MIT as a valuable benchmark for future research. The code is avalibale at: https://github.com/showlab/Multi-human-Talking-Video-Dataset.
title Multi-human Interactive Talking Dataset
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.03050