NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Learned Aggregation of Convolutional Feature Maps
This work addresses the challenge of disentanglement in representation learning for computer vision, but it is incremental as it builds on existing pretrained networks and fine-tuning techniques.
The paper tackled the problem of learning disentangled latent factors in images by proposing a method that aggregates pretrained convolutional feature maps and fine-tunes them on auxiliary tasks, achieving 2nd place in the NeurIPS 2019 disentanglement challenge.
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.