This code contains PyTorch implementation of the paper "Local-Global Graph Pooling via Mutual Information Maximization for Video-Paragraph Retrieval", which has been accepted by IEEE Transactions on Circuits and Systems for Video Technology (TCSVT).
Python 3.6 and PyTorch 1.6.
Install required packages using the environment.yml file.
conda env create -f environment.yml
ActivityNet Captions, Youcook2 with ImageNet/Kinetics Features, and Youcook2 with Howto100m features are used in our experiments.
Our code is based on the implementations of HGR(CVPR2020) and COOT(NeurIPS2020).