Junshen Xu

Results 1 issues of Junshen Xu

Hi, thanks for sharing this fantastic work. I am trying to understand the cross attention used in the model. Here, the conditional context only has one token, i.e., the clip...