Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userAdversarially Robust One-class Novelty Detection
141
0
0.0
(
0
)
Added by
Shao-Yuan Lo
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
One-class novelty detectors are trained with examples of a particular class and are tasked with identifying whether a query example belongs to the same known class. Most recent advances adopt a deep auto-encoder style architecture to compute novelty scores for detecting novel class data. Deep networks have shown to be vulnerable to adversarial attacks, yet little focus is devoted to studying the adversarial robustness of deep novelty detectors. In this paper, we first show that existing novelty detectors are susceptible to adversarial examples. We further demonstrate that commonly-used defense approaches for classification tasks have limited effectiveness in one-class novelty detection. Hence, we need a defense specifically designed for novelty detection. To this end, we propose a defense strategy that manipulates the latent space of novelty detectors to improve the robustness against adversarial examples. The proposed method, referred to as Principal Latent Space (PLS), learns the incrementally-trained cascade principal components in the latent space to robustify novelty detectors. PLS can purify latent space against adversarial examples and constrain latent space to exclusively model the known class distribution. We conduct extensive experiments on multiple attacks, datasets and novelty detectors, showing that PLS consistently enhances the adversarial robustness of novelty detection models.
rate research
Read More
One-class novelty detection is conducted to iden-tify anomalous instances, with different distributions from theexpected normal instances. In this paper, the Generative Adver-sarial Network based on the Encoder-Decoder-Encoder scheme(EDE-GAN) achieves state-of-the-art performance. The two fac-tors bellow serve the above purpose: 1) The EDE-GAN calculatesthe distance between two latent vectors as the anomaly score,which is unlike the previous methods by utilizing the reconstruc-tion error between images. 2) The model obtains best resultswhen the batch size is set to 1. To illustrate their superiority,we design a new GAN architecture, and compareperformances according to different batch sizes. Moreover, withexperimentation leads to discovery, our result implies there is alsoevidence of just how beneficial constraint on the latent space arewhen engaging in model training.In an attempt to learn compact and fast models, we present anew technology, Progressive Knowledge Distillation with GANs(P-KDGAN), which connects two standard GANs through thedesigned distillation loss. Two-step progressive learning continu-ously augments the performance of student GANs with improvedresults over single-step approach. Our experimental results onCIFAR-10, MNIST, and FMNIST datasets illustrate that P-KDGAN improves the performance of the student GAN by2.44%, 1.77%, and 1.73% when compressing the computationat ratios of 24.45:1, 311.11:1, and 700:1, respectively.
We propose a new method for novelty detection that can tolerate high corruption of the training points, whereas previous works assumed either no or very low corruption. Our method trains a robust variational autoencoder (VAE), which aims to generate a model for the uncorrupted training points. To gain robustness to high corruption, we incorporate the following four changes to the common VAE: 1. Extracting crucial features of the latent code by a carefully designed dimension reduction component for distributions; 2. Modeling the latent distribution as a mixture of Gaussian low-rank inliers and full-rank outliers, where the testing only uses the inlier model; 3. Applying the Wasserstein-1 metric for regularization, instead of the Kullback-Leibler (KL) divergence; and 4. Using a least absolute deviation error for reconstruction. We establish both robustness to outliers and suitability to low-rank modeling of the Wasserstein metric as opposed to the KL divergence. We illustrate state-of-the-art results on standard benchmarks for novelty detection.
Neural networks are proven to be remarkably successful for classification and diagnosis in medical applications. However, the ambiguity in the decision-making process and the interpretability of the learned features is a matter of concern. In this work, we propose a method for improving the feature interpretability of neural network classifiers. Initially, we propose a baseline convolutional neural network with state of the art performance in terms of accuracy and weakly supervised localization. Subsequently, the loss is modified to integrate robustness to adversarial examples into the training process. In this work, feature interpretability is quantified via evaluating the weakly supervised localization using the ground truth bounding boxes. Interpretability is also visually assessed using class activation maps and saliency maps. The method is applied to NIH ChestX-ray14, the largest publicly available chest x-rays dataset. We demonstrate that the adversarially robust optimization paradigm improves feature interpretability both quantitatively and visually.
Transfer learning is a widely-used paradigm in deep learning, where models pre-trained on standard datasets can be efficiently adapted to downstream tasks. Typically, better pre-trained models yield better transfer results, suggesting that initial accuracy is a key aspect of transfer learning performance. In this work, we identify another such aspect: we find that adversarially robust models, while less accurate, often perform better than their standard-trained counterparts when used for transfer learning. Specifically, we focus on adversarially robust ImageNet classifiers, and show that they yield improved accuracy on a standard suite of downstream classification tasks. Further analysis uncovers more differences between robust and standard models in the context of transfer learning. Our results are consistent with (and in fact, add to) recent hypotheses stating that robustness leads to improved feature representations. Our code and models are available at https://github.com/Microsoft/robust-models-transfer .
Anomaly detection is a fundamental problem in computer vision area with many real-world applications. Given a wide range of images belonging to the normal class, emerging from some distribution, the objective of this task is to construct the model to detect out-of-distribution images belonging to abnormal instances. Semi-supervised Generative Adversarial Networks (GAN)-based methods have been gaining popularity in anomaly detection task recently. However, the training process of GAN is still unstable and challenging. To solve these issues, a novel adversarial dual autoencoder network is proposed, in which the underlying structure of training data is not only captured in latent feature space, but also can be further restricted in the space of latent representation in a discriminant manner, leading to a more accurate detector. In addition, the auxiliary autoencoder regarded as a discriminator could obtain an more stable training process. Experiments show that our model achieves the state-of-the-art results on MNIST and CIFAR10 datasets as well as GTSRB stop signs dataset.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
