Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Xiangyue, Li, Jianfang, Zhang, Jiaxu, Ren, Jianqiang, Bo, Liefeng, Tu, Zhigang
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2504.09209
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912464289923072
author Zhang, Xiangyue
Li, Jianfang
Zhang, Jiaxu
Ren, Jianqiang
Bo, Liefeng
Tu, Zhigang
author_facet Zhang, Xiangyue
Li, Jianfang
Zhang, Jiaxu
Ren, Jianqiang
Bo, Liefeng
Tu, Zhigang
contents Masked modeling framework has shown promise in co-speech motion generation. However, it struggles to identify semantically significant frames for effective motion masking. In this work, we propose a speech-queried attention-based mask modeling framework for co-speech motion generation. Our key insight is to leverage motion-aligned speech features to guide the masked motion modeling process, selectively masking rhythm-related and semantically expressive motion frames. Specifically, we first propose a motion-audio alignment module (MAM) to construct a latent motion-audio joint space. In this space, both low-level and high-level speech features are projected, enabling motion-aligned speech representation using learnable speech queries. Then, a speech-queried attention mechanism (SQA) is introduced to compute frame-level attention scores through interactions between motion keys and speech queries, guiding selective masking toward motion frames with high attention scores. Finally, the motion-aligned speech features are also injected into the generation network to facilitate co-speech motion generation. Qualitative and quantitative evaluations confirm that our method outperforms existing state-of-the-art approaches, successfully producing high-quality co-speech motion.
format Preprint
id arxiv_https___arxiv_org_abs_2504_09209
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle EchoMask: Speech-Queried Attention-based Mask Modeling for Holistic Co-Speech Motion Generation
Zhang, Xiangyue
Li, Jianfang
Zhang, Jiaxu
Ren, Jianqiang
Bo, Liefeng
Tu, Zhigang
Graphics
Computer Vision and Pattern Recognition
Sound
Masked modeling framework has shown promise in co-speech motion generation. However, it struggles to identify semantically significant frames for effective motion masking. In this work, we propose a speech-queried attention-based mask modeling framework for co-speech motion generation. Our key insight is to leverage motion-aligned speech features to guide the masked motion modeling process, selectively masking rhythm-related and semantically expressive motion frames. Specifically, we first propose a motion-audio alignment module (MAM) to construct a latent motion-audio joint space. In this space, both low-level and high-level speech features are projected, enabling motion-aligned speech representation using learnable speech queries. Then, a speech-queried attention mechanism (SQA) is introduced to compute frame-level attention scores through interactions between motion keys and speech queries, guiding selective masking toward motion frames with high attention scores. Finally, the motion-aligned speech features are also injected into the generation network to facilitate co-speech motion generation. Qualitative and quantitative evaluations confirm that our method outperforms existing state-of-the-art approaches, successfully producing high-quality co-speech motion.
title EchoMask: Speech-Queried Attention-based Mask Modeling for Holistic Co-Speech Motion Generation
topic Graphics
Computer Vision and Pattern Recognition
Sound
url https://arxiv.org/abs/2504.09209