CaltechAUTHORS
  A Caltech Library Service

Improving the Robustness of Deep Neural Networks via Stability Training

Zheng, Stephan and Song, Yang and Leung, Thomas and Goodfellow, Ian (2016) Improving the Robustness of Deep Neural Networks via Stability Training. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE , Piscataway, NJ, pp. 4480-4488. ISBN 978-1-4673-8851-1. http://resolver.caltech.edu/CaltechAUTHORS:20170525-102807503

Full text is not posted in this repository. Consult Related URLs below.

Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechAUTHORS:20170525-102807503

Abstract

In this paper we address the issue of output instability of deep neural networks: small perturbations in the visual input can significantly distort the feature embeddings and output of a neural network. Such instability affects many deep architectures with state-of-the-art performance on a wide range of computer vision tasks. We present a general stability training method to stabilize deep networks against small input distortions that result from various types of common image processing, such as compression, rescaling, and cropping. We validate our method by stabilizing the state of-the-art Inception architecture [11] against these types of distortions. In addition, we demonstrate that our stabilized model gives robust state-of-the-art performance on largescale near-duplicate detection, similar-image ranking, and classification on noisy datasets.


Item Type:Book Section
Related URLs:
URLURL TypeDescription
http://dx.doi.org/10.1109/CVPR.2016.485DOIArticle
http://ieeexplore.ieee.org/document/7780854/PublisherArticle
Additional Information:© 2017 IEEE.
Subject Keywords:Training, Stability analysis, Robustness, Neural networks, Visualization, Feature extraction, Data models
Record Number:CaltechAUTHORS:20170525-102807503
Persistent URL:http://resolver.caltech.edu/CaltechAUTHORS:20170525-102807503
Official Citation:S. Zheng, Y. Song, T. Leung and I. Goodfellow, "Improving the Robustness of Deep Neural Networks via Stability Training," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, 2016, pp. 4480-4488. doi: 10.1109/CVPR.2016.485 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7780854&isnumber=7780329
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:77754
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:25 May 2017 18:05
Last Modified:25 May 2017 18:05

Repository Staff Only: item control page