Gradient-based Training of Slow Feature Analysis by Differentiable Approximate Whitening

Merlin Schüler, Hlynur Davíð Hlynsson, Laurenz Wiskott
Proceedings of The Eleventh Asian Conference on Machine Learning, PMLR 101:316-331, 2019.

Abstract

We propose Power Slow Feature Analysis, a gradient-based method to extract temporally slow features from a high-dimensional input stream that varies on a faster time-scale, as a variant of Slow Feature Analysis (SFA) that allows end-to-end training of arbitrary differentiable architectures and thereby significantly extends the class of models that can effectively be used for slow feature extraction. We provide experimental evidence that PowerSFA is able to extract meaningful and informative low-dimensional features in the case of (a) synthetic low-dimensional data, (b) ego-visual data, and also for (c) a general dataset for which symmetric non-temporal similarities between points can be defined.

Cite this Paper

BibTeX
@InProceedings{pmlr-v101-schuler19a, title = {Gradient-based Training of Slow Feature Analysis by Differentiable Approximate Whitening}, author = {Sch{\"u}ler, Merlin and Hlynsson, Hlynur Dav\'i\dh and Wiskott, Laurenz}, booktitle = {Proceedings of The Eleventh Asian Conference on Machine Learning}, pages = {316--331}, year = {2019}, editor = {Lee, Wee Sun and Suzuki, Taiji}, volume = {101}, series = {Proceedings of Machine Learning Research}, month = {17--19 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v101/schuler19a/schuler19a.pdf}, url = {https://proceedings.mlr.press/v101/schuler19a.html}, abstract = {We propose Power Slow Feature Analysis, a gradient-based method to extract temporally slow features from a high-dimensional input stream that varies on a faster time-scale, as a variant of Slow Feature Analysis (SFA) that allows end-to-end training of arbitrary differentiable architectures and thereby significantly extends the class of models that can effectively be used for slow feature extraction. We provide experimental evidence that PowerSFA is able to extract meaningful and informative low-dimensional features in the case of (a) synthetic low-dimensional data, (b) ego-visual data, and also for (c) a general dataset for which symmetric non-temporal similarities between points can be defined. } }
Endnote
%0 Conference Paper %T Gradient-based Training of Slow Feature Analysis by Differentiable Approximate Whitening %A Merlin Schüler %A Hlynur Davíð Hlynsson %A Laurenz Wiskott %B Proceedings of The Eleventh Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Wee Sun Lee %E Taiji Suzuki %F pmlr-v101-schuler19a %I PMLR %P 316--331 %U https://proceedings.mlr.press/v101/schuler19a.html %V 101 %X We propose Power Slow Feature Analysis, a gradient-based method to extract temporally slow features from a high-dimensional input stream that varies on a faster time-scale, as a variant of Slow Feature Analysis (SFA) that allows end-to-end training of arbitrary differentiable architectures and thereby significantly extends the class of models that can effectively be used for slow feature extraction. We provide experimental evidence that PowerSFA is able to extract meaningful and informative low-dimensional features in the case of (a) synthetic low-dimensional data, (b) ego-visual data, and also for (c) a general dataset for which symmetric non-temporal similarities between points can be defined.
APA
Schüler, M., Hlynsson, H.D. & Wiskott, L.. (2019). Gradient-based Training of Slow Feature Analysis by Differentiable Approximate Whitening. Proceedings of The Eleventh Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 101:316-331 Available from https://proceedings.mlr.press/v101/schuler19a.html.

Related Material