Songfang Huang

A new method enhances efficiency and performance in vision-language tasks.

A new method enhances event resolution by combining language models for better accuracy.

Advancing task-solving models for languages with limited data through innovative merging techniques.