Maximum-Roaming-Mutli-Task-Learning

This code was made under Pytorch v1.2.0, and is associated to the paper https://arxiv.org/pdf/2006.09762.pdf.

Each provided folder contains code for its respective dataset (Celeb-A, CityScapes and NYUv2). Inside of these, individual training scripts are provided for each of the baselines described in the submission paper.

For the Celeb-A dataset, we use the official release : http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html

For CityScapes and NYUv2, we use the preprocessed data made available here : https://github.com/lorenmt/mtan

How to cite

Please cite as follows if you use this work:

Pascal, L., Michiardi, P., Bost, X.,Hüet, B., Zuluaga, M. A. (2021).
Maximum Roaming Multi-task Learning
Proceedings of the 35th AAAI International Conference on Artificial Intelligence, February 2-9, 2021 (in press)
arXiv preprint arXiv:2006.09762