VAL icon indicating copy to clipboard operation
VAL copied to clipboard

Where is the code of Auxiliary visual-semantic matching?

Open huynhtruc0309 opened this issue 3 years ago • 1 comments

Hello. I cannot find the code of the Lvs loss. Please help me. Thank you. image

huynhtruc0309 avatar Mar 31 '21 03:03 huynhtruc0309

Hi, This part can be done by pre-training the base network using the same loss, but the different texts (i.e. tagged captions in fasion200k). So the difference is only to replace the text input.

yanbeic avatar Apr 26 '21 09:04 yanbeic