/STL-VQA

The good practice in the VQA system such as pos-tag attention, structed triplet learning and triplet attention is very general and can be inserted into almost any visual and language task

Primary LanguagePython

Watchers