HumanGuidedAttention

The code for "Human-Guided Modality Informativeness for Affective States" at ICMI 2021.

Installation

Python 3.10 and poetry are required.

poetry add git+https://github.com:twoertwein/HumanGuidedAttention.git

To locate human_guided_attention.train, please use

poetry run python -c "from human_guided_attention import train; print(train.__file__)"

Usage

Run the grid-search for the guided model:

poetry run python path/to/train.py --GUIDED

Citation and Dataset

The creation of the Transitions in Parenting of Teens (TPOT) dataset was funded by NIH grant #5R01 HD081362 (awarded to Lisa B. Sheeber and Nicholas B. Allen). When referring to the TPOT dataset, please cite

@article{nelson2021psychobiological,
  title={Psychobiological markers of allostatic load in depressed and nondepressed mothers and their adolescent offspring},
  author={Nelson, Benjamin W and Sheeber, Lisa and Pfeifer, Jennifer and Allen, Nicholas B},
  journal={Journal of Child Psychology and Psychiatry},
  volume={62},
  number={2},
  pages={199--211},
  year={2021},
  publisher={Wiley Online Library}
}

When using the multimodal features and annotations collected as part of this ICMI paper, please cite

@inproceedings{wortwein2021human,
  title={Human-Guided Modality Informativeness for Affective States},
  author={W{\"o}rtwein, Torsten and Sheeber, Lisa B and Allen, Nicholas and Cohn, Jeffrey F and Morency, Louis-Philippe},
  booktitle={Proceedings of the 2021 International Conference on Multimodal Interaction},
  pages={728--734},
  year={2021}
}