Earlier Version

KS(conf): A light-weight test if a ConvNet operates outside of Its specifications

R. Sun, C. Lampert, in:, Springer Nature, 2019, pp. 244–259.


Conference Paper | Published | English
Author
Department
Series Title
LNCS
Abstract
Computer vision systems for automatic image categorization have become accurate and reliable enough that they can run continuously for days or even years as components of real-world commercial applications. A major open problem in this context, however, is quality control. Good classification performance can only be expected if systems run under the specific conditions, in particular data distributions, that they were trained for. Surprisingly, none of the currently used deep network architectures have a built-in functionality that could detect if a network operates on data from a distribution it was not trained for, such that potentially a warning to the human users could be triggered. In this work, we describe KS(conf), a procedure for detecting such outside of specifications (out-of-specs) operation, based on statistical testing of the network outputs. We show by extensive experiments using the ImageNet, AwA2 and DAVIS datasets on a variety of ConvNets architectures that KS(conf) reliably detects out-of-specs situations. It furthermore has a number of properties that make it a promising candidate for practical deployment: it is easy to implement, adds almost no overhead to the system, works with all networks, including pretrained ones, and requires no a priori knowledge of how the data distribution could change.
Publishing Year
Date Published
2019-02-14
Volume
11269
Page
244-259
Conference
GCPR: Conference on Pattern Recognition
Conference Location
Stuttgart, Germany
Conference Date
2018-10-09 – 2018-10-12
IST-REx-ID

Cite this

Sun R, Lampert C. KS(conf): A light-weight test if a ConvNet operates outside of Its specifications. In: Vol 11269. Springer Nature; 2019:244-259. doi:10.1007/978-3-030-12939-2_18
Sun, R., & Lampert, C. (2019). KS(conf): A light-weight test if a ConvNet operates outside of Its specifications (Vol. 11269, pp. 244–259). Presented at the GCPR: Conference on Pattern Recognition, Stuttgart, Germany: Springer Nature. https://doi.org/10.1007/978-3-030-12939-2_18
Sun, Rémy, and Christoph Lampert. “KS(Conf): A Light-Weight Test If a ConvNet Operates Outside of Its Specifications,” 11269:244–59. Springer Nature, 2019. https://doi.org/10.1007/978-3-030-12939-2_18.
R. Sun and C. Lampert, “KS(conf): A light-weight test if a ConvNet operates outside of Its specifications,” presented at the GCPR: Conference on Pattern Recognition, Stuttgart, Germany, 2019, vol. 11269, pp. 244–259.
Sun R, Lampert C. 2019. KS(conf): A light-weight test if a ConvNet operates outside of Its specifications. GCPR: Conference on Pattern Recognition, LNCS, vol. 11269. 244–259.
Sun, Rémy, and Christoph Lampert. KS(Conf): A Light-Weight Test If a ConvNet Operates Outside of Its Specifications. Vol. 11269, Springer Nature, 2019, pp. 244–59, doi:10.1007/978-3-030-12939-2_18.

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data IST Research Explorer

Sources

arXiv 1804.04171

Search this title in

Google Scholar
ISBN Search