Abstract
We investigate the performance of DNNs when trained on class-incremental visual problems consisting of initial training, followed by retraining with added visual classes. Catastrophic forgetting (CF) behavior is measured using a new evaluation procedure that aims at an application-oriented view of incremental learning. In particular, it imposes that model selection must be performed on the initial dataset alone, as well as demanding that retraining control be performed only using the retraining dataset, as initial dataset is usually too large to be kept. Experiments are conducted on class-incremental problems derived from MNIST, using a variety of different DNN models, some of them recently proposed to avoid catastrophic forgetting. When comparing our new evaluation procedure to previous approaches for assessing CF, we find their findings are completely negated, and that none of the tested methods can avoid CF in all experiments. This stresses the importance of a realistic empirical measurement procedure for catastrophic forgetting, and the need for further research in incremental learning for DNNs.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ciresan, D.C., Meier, U., Masci, J., Maria Gambardella, L., Schmidhuber, J.: Flexible, high performance convolutional neural networks for image classification. In: IJCAI Proceedings of International Joint Conference on Artificial Intelligence, Barcelona, Spain, vol. 22, p. 1237 (2011)
French, R.: Catastrophic forgetting in connectionist networks. Trends Cogn. Sci. (4) (1999)
Goodfellow, I.J., Mirza, M., Xiao, D., Courville, A., Bengio, Y.: An empirical investigation of catastrophic forgetting in gradient-based neural networks. arXiv preprint arXiv:1312.6211 (2013)
Hinton, G.E., Srivastava, N., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.R.: Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv:1207.0580 (2012)
Kirkpatrick, J., et al.: Overcoming catastrophic forgetting in neural networks. Proc. Natl. Acad. Sci. 114, 3521–3526 (2017). https://doi.org/10.1073/pnas.1611835114
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. In: Intelligent Signal Processing. IEEE Press (2001)
Lee, S.W., Kim, J.H., Jun, J., Ha, J.W., Zhang, B.T.: Overcoming catastrophic forgetting by incremental moment matching. In: Advances in Neural Information Processing Systems, pp. 4655–4665 (2017)
Mnih, V., et al.: Human-level control through deep reinforcement learning. Nature 7540, 529–533 (2015)
Sigaud, O., Salaün, C., Padois, V.: On-line regression algorithms for learning mechanical models of robots: a survey. Rob. Auton. Syst. 12, 1115–1129 (2011)
Srivastava, R.K., Masci, J., Kazerounian, S., Gomez, F., Schmidhuber, J.: Compete to compute. In: Advances in Neural Information Processing Systems (2013)
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this paper
Cite this paper
Pfülb, B., Gepperth, A., Abdullah, S., Kilian, A. (2018). Catastrophic Forgetting: Still a Problem for DNNs. In: Kůrková, V., Manolopoulos, Y., Hammer, B., Iliadis, L., Maglogiannis, I. (eds) Artificial Neural Networks and Machine Learning – ICANN 2018. ICANN 2018. Lecture Notes in Computer Science(), vol 11139. Springer, Cham. https://doi.org/10.1007/978-3-030-01418-6_48
Download citation
DOI: https://doi.org/10.1007/978-3-030-01418-6_48
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-01417-9
Online ISBN: 978-3-030-01418-6
eBook Packages: Computer ScienceComputer Science (R0)