VideoLanguageFuturePrediction

VLEP dataset for video and language future event prediction.

What is More Likely to Happen Next? Video-and-Language Future Event Prediction

Jie Lei, Licheng Yu, Tamara L. Berg, Mohit Bansal

Table of Contents

• VideoLanguageFuturePrediction
• Dataset
• Evaluation and CodaLab Submission
• Related work
• Citation
• Contact

VLEP Dataset

The dataset is released at data, please see data/README.md for details.

Evaluation and CodaLab Submission

We only release ground-truth answers for train and dev splits. To get results on the test split, please submit your results to our CodaLab evaluation server following the instructions here: standalone_eval/README.md.

Related Work

• TVC (Video+Dialogue Captioning)
• TVR (Video+Dialogue Retrieval)
• TVQA (Localized Video QA)
• TVQA+ (Spatio-Temporal Video QA)
• recurrent-transformer (coherent video paragraph captioning)

Citation

If you find this code useful for your research, please cite our paper:

@inproceedings{lei2020vlep,
  title={What is More Likely to Happen Next? Video-and-Language Future Event Prediction},
  author={Lei, Jie and Yu, Licheng and Berg, Tamara L and Bansal, Mohit},
  booktitle={EMNLP},
  year={2020}
}

Contact

Jie Lei, jielei@cs.unc.edu