about “Preprocessed caption graphs and trained embeddings”

usydnlp / VICTR

This repository contains code for paper VICTR: Visual Information Captured Text Representation for Text-to-Image Multimodal Tasks

13 stars 4 forks source link

about “Preprocessed caption graphs and trained embeddings” #1

Closed JackWhite-rwx closed 3 years ago

JackWhite-rwx commented 3 years ago

@usydnlp Thanks once again for awesome work.this work is so cool，it inspired me a lot.but when I try to learn this code.I don't find 'text to scene graph ' part.Could you help me? Thank you very much.