Jul, 2024
M$^2$IST: 多模式交互侧调节用于记忆效率的指称表达理解
M$^2$IST: Multi-Modal Interactive Side-Tuning for Memory-efficient Referring Expression Comprehension
Xuyang Liu, Ting Liu, Siteng Huang, Yue Hu, Quanjun Yin...
TL;DRReferring expression comprehension is improved through M$^2$IST, a parameter- and memory-efficient transfer learning method utilizing M$^3$ISAs for establishing connections between pre-trained vision and language encoders.