[Paper Notes] BEIT V2: Masked Image Modeling with Vector-Quantized Visual Tokenizers

NoSuchKey

Guess you like

Origin blog.csdn.net/weixin_50862344/article/details/131262830