LoSATok: Low-dimensional Semantic-Acoustic Tokenizer for Cross-Domain Audio Understanding and Generation 文章

ArXiv CS.AI2026-05-28NEWSen作者: Zhisheng Zhang, Xiang Li, Yixuan Zhou, Jing Peng, Guoyang Zeng, Zhiyong Wu
查看原文 →
关系图谱

LoSATok: Low-dimensional Semantic-Acoustic Tokenizer for Cross-Domain Audio Understanding and Generation · 相关技术

相关技术

dual-level semantic supervisiontime-relation lossSemantic BottleneckDiffusion Transformer