Open czhxiaohuihui opened 4 years ago
First you should train Large-Scale-VRD model with GQA dataset, then you can get the relation embedding via removing semantic module in Large-Scale-VRD model.
More detail in https://github.com/jz462/Large-Scale-VRD.pytorch
Can you explain how to use Large-Scale-VRD to extract visual relation embedding exactly?Very Thanks!!!!