This repo corresponds to the submission 'Hierarchical Multi-Source Uncertainty Aggregation for Interactive Video Captioning'.

The repo contains the following scripts:

main.py : The main function for the interactive learning process.

preprocess.py : The script for data preprocessing.

nltkeval.py : The script containing functions for evaluation.

utilInterpret.py : The script containing functions for loading and interpretation.

utilVisualize.py : The script containing functions for visualization.

model.py : The script containing the model.

To get started, download dataset from https://www.kaggle.com/datasets/vishnutheepb/msrvtt/data and run preprocess.py.

After that, run main.py for the interactive learning process.