Guardado en:
Detalles Bibliográficos
Autores principales: Guo, Ying, Gan, Qijun, Zhang, Yifu, Liu, Jinlai, Hu, Yifei, Xie, Pan, Qian, Dongjun, Zhang, Yu, Li, Ruiqi, Zhang, Yuqi, Lu, Ruibiao, Mei, Xiaofeng, Han, Bo, Yin, Xiang, Peng, Bingyue, Yuan, Zehuan
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2602.08682
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Tabla de Contenidos:
  • Video generation is rapidly evolving towards unified audio-video generation. In this paper, we present ALIVE, a generation model that adapts a pretrained Text-to-Video (T2V) model to Sora-style audio-video generation and animation. In particular, the model unlocks the Text-to-Video&Audio (T2VA) and Reference-to-Video&Audio (animation) capabilities compared to the T2V foundation models. To support the audio-visual synchronization and reference animation, we augment the popular MMDiT architecture with a joint audio-video branch which includes TA-CrossAttn for temporally-aligned cross-modal fusion and UniTemp-RoPE for precise audio-visual alignment. Meanwhile, a comprehensive data pipeline consisting of audio-video captioning, quality control, etc., is carefully designed to collect high-quality finetuning data. Additionally, we introduce a new benchmark to perform a comprehensive model test and comparison. After continue pretraining and finetuning on million-level high-quality data, ALIVE demonstrates outstanding performance, consistently outperforming open-source models and matching or surpassing state-of-the-art commercial solutions. With detailed recipes and benchmarks, we hope ALIVE helps the community develop audio-video generation models more efficiently. Official page: https://github.com/FoundationVision/Alive.