1 paper · 1 filter
Lirong Che, Zhenfeng Gan, Yanbo Chen +2
Embodied agents for creative tasks like photography must bridge the semantic gap between high-level language commands and geometric control. We introduce PhotoAgent, an agent that…