Learning by association a versatile semi-supervised training method for neural networks
Haeusser P, Mordvintsev A, Cremers D (2017)
Publication Type: Conference contribution
Publication year: 2017
Publisher: Institute of Electrical and Electronics Engineers Inc.
Book Volume: 2017-January
Pages Range: 626-635
Conference Proceedings Title: Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017
Event location: Honolulu, HI, USA
ISBN: 9781538604571
DOI: 10.1109/CVPR.2017.74
Abstract
In many real-world scenarios, labeled data for a specific machine learning task is costly to obtain. Semi-supervised training methods make use of abundantly available unlabeled data and a smaller number of labeled examples. We propose a new framework for semi-supervised training of deep neural networks inspired by learning in humans. “Associations” are made from embeddings of labeled samples to those of unlabeled ones and back. The optimization schedule encourages correct association cycles that end up at the same class from which the association was started and penalizes wrong associations ending at a different class. The implementation is easy to use and can be added to any existing end-to-end training setup. We demonstrate the capabilities of learning by association on several data sets and show that it can improve performance on classification tasks tremendously by making use of additionally available unlabeled data. In particular, for cases with few labeled data, our training scheme outperforms the current state of the art on SVHN.
Involved external institutions
Germany (DE)
United States (USA) (US)How to cite
APA:
Haeusser, P., Mordvintsev, A., & Cremers, D. (2017). Learning by association a versatile semi-supervised training method for neural networks. In Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017 (pp. 626-635). Honolulu, HI, USA: Institute of Electrical and Electronics Engineers Inc..
MLA:
Haeusser, Philip, Alexander Mordvintsev, and Daniel Cremers. "Learning by association a versatile semi-supervised training method for neural networks." Proceedings of the 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017, Honolulu, HI, USA Institute of Electrical and Electronics Engineers Inc., 2017. 626-635.
BibTeX: Download