Guardado en:
Detalles Bibliográficos
Autores principales: Saleem, Muhammad Usama, Patel, Mayur Jagdishbhai, Pinyoanuntapong, Ekkasit, Qin, Zhongxing, Yang, Li, Xue, Hongfei, Helmy, Ahmed, Chen, Chen, Wang, Pu
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2604.10927
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911586884517888
author Saleem, Muhammad Usama
Patel, Mayur Jagdishbhai
Pinyoanuntapong, Ekkasit
Qin, Zhongxing
Yang, Li
Xue, Hongfei
Helmy, Ahmed
Chen, Chen
Wang, Pu
author_facet Saleem, Muhammad Usama
Patel, Mayur Jagdishbhai
Pinyoanuntapong, Ekkasit
Qin, Zhongxing
Yang, Li
Xue, Hongfei
Helmy, Ahmed
Chen, Chen
Wang, Pu
contents We propose LiveGesture, the first fully streamable, speech-driven full-body gesture generation framework that operates with zero look-ahead and supports arbitrary sequence length. Unlike existing co-speech gesture methods, which are designed for offline generation and either treat body regions independently or entangle all joints within a single model, LiveGesture is built from the ground up for causal, region-coordinated motion generation. LiveGesture consists of two main modules: the Streamable Vector Quantized Motion Tokenizer (SVQ) and the Hierarchical Autoregressive Transformer (HAR). The SVQ tokenizer converts the motion sequence of each body region into causal, discrete motion tokens, enabling real-time, streamable token decoding. On top of SVQ, HAR employs region-expert autoregressive (xAR) transformers to model expressive, fine-grained motion dynamics for each body region. A causal spatio-temporal fusion module (xAR Fusion) then captures and integrates correlated motion dynamics across regions. Both xAR and xAR Fusion are conditioned on live, continuously arriving audio signals encoded by a streamable causal audio encoder. To enhance robustness under streaming noise and prediction errors, we introduce autoregressive masking training, which leverages uncertainty-guided token masking and random region masking to expose the model to imperfect, partially erroneous histories during training. Experiments on the BEAT2 dataset demonstrate that LiveGesture produces coherent, diverse, and beat-synchronous full-body gestures in real time, matching or surpassing state-of-the-art offline methods under true zero look-ahead conditions.
format Preprint
id arxiv_https___arxiv_org_abs_2604_10927
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LiveGesture Streamable Co-Speech Gesture Generation Model
Saleem, Muhammad Usama
Patel, Mayur Jagdishbhai
Pinyoanuntapong, Ekkasit
Qin, Zhongxing
Yang, Li
Xue, Hongfei
Helmy, Ahmed
Chen, Chen
Wang, Pu
Computer Vision and Pattern Recognition
We propose LiveGesture, the first fully streamable, speech-driven full-body gesture generation framework that operates with zero look-ahead and supports arbitrary sequence length. Unlike existing co-speech gesture methods, which are designed for offline generation and either treat body regions independently or entangle all joints within a single model, LiveGesture is built from the ground up for causal, region-coordinated motion generation. LiveGesture consists of two main modules: the Streamable Vector Quantized Motion Tokenizer (SVQ) and the Hierarchical Autoregressive Transformer (HAR). The SVQ tokenizer converts the motion sequence of each body region into causal, discrete motion tokens, enabling real-time, streamable token decoding. On top of SVQ, HAR employs region-expert autoregressive (xAR) transformers to model expressive, fine-grained motion dynamics for each body region. A causal spatio-temporal fusion module (xAR Fusion) then captures and integrates correlated motion dynamics across regions. Both xAR and xAR Fusion are conditioned on live, continuously arriving audio signals encoded by a streamable causal audio encoder. To enhance robustness under streaming noise and prediction errors, we introduce autoregressive masking training, which leverages uncertainty-guided token masking and random region masking to expose the model to imperfect, partially erroneous histories during training. Experiments on the BEAT2 dataset demonstrate that LiveGesture produces coherent, diverse, and beat-synchronous full-body gestures in real time, matching or surpassing state-of-the-art offline methods under true zero look-ahead conditions.
title LiveGesture Streamable Co-Speech Gesture Generation Model
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.10927