Si Liu

MARBLE sets a standard for evaluating music AI models across multiple tasks.

LyricWhiz combines advanced models to improve lyric transcription accuracy across languages.

DiffDance creates detailed dance sequences that match music effectively.

A new method aids robots in carrying objects collaboratively.

LLaVA-MoD creates smaller multimodal models using knowledge from larger counterparts.

Discover how images are created through advanced technologies and models.

Learn how database transactions ensure data consistency and efficiency.

A new dataset enhancing video understanding and AI reasoning.