Xia Zhiqi

Results 1 comments of Xia Zhiqi

> Hi @Mountchicken ,I have some questions in some Model Details, Trex2 paper say they use CLIP as text encoder and use [CLS] token to get the text embedding. If...