Text with Knowledge Graph Augmented Transformer for Video Captioning

Introduction

This repository provides the code and resources associated with our paper titled Text with Knowledge Graph Augmented Transformer for Video Captioning, which has been accepted by CVPR 2023. The paper presents a novel approach that leverages knowledge graphs to deal with the the challenge of long-tail words in video captioning tasks.

[paper]

[The code has been released here~]

We plan to release the code as open source in mid-June. Stay tuned to our GitHub repository for the latest updates and notifications.