Multimodal Joint Emotion and Game Context Recognition in League of Legends Livestreams

This repository contains the code to support the paper "Multimodal Joint Emotion and Game Context Recognition in League of Legends Livestreams" presented at the 2019 IEEE Conference on Games.

Data

The associated dataset can be downloaded from: https://tinyurl.com/y5qakf2r

Copy all files/folders found in the link to the code root folder if you wish to use them together.

Usage

• The data set needs to be converted into npy files first. This can be done with utils/process_data_set.py.
• (optional) To apply the oversampling technique detailed in the paper run utils/augment_annotations.py.
• Once these preprocessing steps are complete trainer.py will run the full set of experiments.

Requirements

That project requires Python 3.6.3 (other versions may work but are untested). Requirements can be found in requirements.txt. These can be installed with pip install -r requirements.txt.

Cite this work

Please cite this work as: Ringer, Charles, Walker, James Alfred, Nicolaou, Mihalis A. (2019) Multimodal Joint Emotion and Game Context Recognition in League of Legends Livestreams, In: Proceedings of the IEEE Conference on Games 2019. IEEE

Licence

MIT