Open-Set Recognition with Gaussian Mixture Variational Autoencoders
Abstract
In inference, open-set classification is to either classify a sample into a known class from training or reject it as an unknown class. Existing deep open-set classifiers train explicit closed-set classifiers, in some cases disjointly utilizing reconstruction, which we find dilutes the latent representation's ability to distinguish unknown classes. In contrast, we train our model to cooperatively learn reconstruction and perform class-based clustering in the latent space. With this, our Gaussian mixture variational autoencoder (GMVAE) achieves more accurate and robust open-set classification results, with an average F1 increase of 0.26, through extensive experiments aided by analytical results.
Cite
Text
Cao et al. "Open-Set Recognition with Gaussian Mixture Variational Autoencoders." AAAI Conference on Artificial Intelligence, 2021. doi:10.1609/AAAI.V35I8.16848Markdown
[Cao et al. "Open-Set Recognition with Gaussian Mixture Variational Autoencoders." AAAI Conference on Artificial Intelligence, 2021.](https://mlanthology.org/aaai/2021/cao2021aaai-open/) doi:10.1609/AAAI.V35I8.16848BibTeX
@inproceedings{cao2021aaai-open,
title = {{Open-Set Recognition with Gaussian Mixture Variational Autoencoders}},
author = {Cao, Alexander and Luo, Yuan and Klabjan, Diego},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2021},
pages = {6877-6884},
doi = {10.1609/AAAI.V35I8.16848},
url = {https://mlanthology.org/aaai/2021/cao2021aaai-open/}
}