NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Learned Aggregation of Convolutional Feature Maps
Abstract
This report to our stage 2 submission to the NeurIPS 2019 disentanglement challenge presents a simple image preprocessing method for learning disentangled latent factors. We propose to train a variational autoencoder on regionally aggregated feature maps obtained from networks pretrained on the ImageNet database, utilizing the implicit inductive bias contained in those features for disentanglement. This bias can be further enhanced by explicitly fine-tuning the feature maps on auxiliary tasks useful for the challenge, such as angle, position estimation, or color classification. Our approach achieved the 2nd place in stage 2 of the challenge. Code is available at https://github.com/mseitzer/neurips2019-disentanglement-challenge.
Keywords
Cite
@article{arxiv.2002.12356,
title = {NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Learned Aggregation of Convolutional Feature Maps},
author = {Maximilian Seitzer and Andreas Foltyn and Felix P. Kemeth},
journal= {arXiv preprint arXiv:2002.12356},
year = {2020}
}
Comments
Disentanglement Challenge - 33rd Conference on Neural Information Processing Systems (NeurIPS) - NeurIPS 2019. arXiv admin note: text overlap with arXiv:2002.10003. Acknowledgements added