This paper proposes to use a frozen vision and language pre-trained model ... (LSJ) recipe (Ghiasi et al 2021), which is also in Detectron2.
確定! 回上一頁