检索结果-内蒙古大学图书馆

Understanding and stabilizing gans' training dynamics with control theory

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Xu, Kun Li, Chongxuan Wei, Huanshu Zhu, Jun Zhang, Bo Department of Computer Science & Technology Institute for Artificial Intelligence BNRist Center THBI Lab State Key Lab for Intell. Tech. & Sys. Tsinghua University

Generative adversarial networks (GANs) have made significant progress on realistic image generation but often suffer from instability during the training process. Most previous analyses mainly focus on the equilibrium that GANs achieve, whereas a gap exists between such theoretical analyses and practical implementations, where it is the training dynamics that plays a vital role in the convergence and stability of GANs. In this paper, we directly model the dynamics of GANs and adopt the control theory to understand and stabilize it. Specifically, we interpret the training process of various GANs as certain types of dynamics in a unified perspective of control theory which enables us to model the stability and convergence easily. Borrowed from control theory, we adopt the widely-used negative feedback control to stabilize the training dynamics, which can be considered as an L2 regularization on the output of the discriminator. We empirically verify our method on both synthetic data and natural image datasets. The results demonstrate that our method can stabilize the training dynamics as well as converge better than baselines. Copyright © 2019, The Authors. All rights reserved.

关键词： Generative adversarial networks

Benchmarking adversarial robustness

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Dong, Yinpeng Fu, Qi-An Yang, Xiao Pang, Tianyu Su, Hang Xiao, Zihao Zhu, Jun Dept. of Comp. Sci. and Tech. BNRist Center State Key Lab for Intell. Tech. & Sys. Institute for AI THBI Lab Tsinghua University Beijing100084 China RealAI

Deep neural networks are vulnerable to adversarial examples, which becomes one of the most important research problems in the development of deep learning. While a lot of efforts have been made in recent years, it is of great significance to perform correct and complete evaluations of the adversarial attack and defense algorithms. In this paper, we establish a comprehensive, rigorous, and coherent benchmark to evaluate adversarial robustness on image classification tasks. After briefly reviewing plenty of representative attack and defense methods, we perform large-scale experiments with two robustness curves as the fair-minded evaluation criteria to fully understand the performance of these methods. Based on the evaluation results, we draw several important findings and provide insights for future research. Copyright © 2019, The Authors. All rights reserved.

关键词： Network security

To relieve your headache of training an MRF, take AdVIL

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Li, Chongxuan Du, Chao Xu, Kun Welling, Max Zhu, Jun Zhang, Bo Department of Computer Science & Technology Institute for Artificial Intelligence BNRist Center THBI Lab State Key Lab for Intell. Tech. & Sys. Tsinghua University

We propose a black-box algorithm called Adversarial Variational Inference and Learning (AdVIL) to perform inference and learning on a general Markov random field (MRF). AdVIL employs two variational distributions to approximately infer the latent variables and estimate the partition function of an MRF, respectively. The two variational distributions provide an estimate of the negative log-likelihood of the MRF as a minimax optimization problem, which is solved by stochastic gradient descent. AdVIL is proven convergent under certain conditions. On one hand, compared with contrastive divergence, AdVIL requires a minimal assumption about the model structure and can deal with a broader family of MRFs. On the other hand, compared with existing black-box methods, AdVIL provides a tighter estimate of the log partition function and achieves much better empirical results. Copyright © 2019, The Authors. All rights reserved.

关键词： Markov processes

DS3L: Deep self-semi-supervised learning for image recognition

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Tsai, Tsung Wei Li, Chongxuan Zhu, Jun Dept. of Comp. Sci. & Tech. BNRist Center State Key Lab for Intell. Tech.& Sys. THBI Lab Tsinghua University Beijing100084 China

Despite the recent progress in deep semi-supervised learning (Semi-SL), the amount of labels still plays a dominant role. The success in self-supervised learning (Self-SL) hints a promising direction to exploit the vast unlabeled data by leveraging an additional set of deterministic labels. In this paper, we propose Deep Self-Semi-Supervised learning (DS3L), a flexible multi-task framework with shared parameters that integrates the rotation task in Self-SL with the consistency-based methods in deep Semi-SL. Our method is easy to implement and is complementary to all consistency-based approaches. The experiments demonstrate that our method significantly improves over the published state-of-the-art methods on several standard benchmarks, especially when fewer labels are presented. Copyright © 2019, The Authors. All rights reserved.

关键词： Supervised learning

Efficient Decision-based Black-box Adversarial Attacks on Face Recognition

学校读者我要写书评

暂无评论

Efficient Decision-based Black-box Adversarial Attacks on Fa...

IEEE/CVF Conference on Computer Vision and Pattern Recognition

作者： Yinpeng Dong Hang Su Baoyuan Wu Zhifeng Li Wei Liu Tong Zhang Jun Zhu Dept. of Comp. Sci. and Tech. BNRist Center State Key Lab for Intell. Tech. & Sys. Institute for AI THBI Lab Tsinghua University Tencent AI Lab Hong Kong University of Science and Technology

ISBN: (纸本)9781728132945

Face recognition has obtained remarkable progress in recent years due to the great improvement of deep convolutional neural networks (CNNs). However, deep CNNs are vulnerable to adversarial examples, which can cause fateful consequences in real-world face recognition applications with security-sensitive purposes. Adversarial attacks are widely studied as they can identify the vulnerability of the models before they are deployed. In this paper, we evaluate the robustness of state-of-the-art face recognition models in the decision-based black-box attack setting, where the attackers have no access to the model parameters and gradients, but can only acquire hard-label predictions by sending queries to the target model. This attack setting is more practical in real-world face recognition systems. To improve the efficiency of previous methods, we propose an evolutionary attack algorithm, which can model the local geometry of the search directions and reduce the dimension of the search space. Extensive experiments demonstrate the effectiveness of the proposed method that induces a minimum perturbation to an input face image with fewer queries. We also apply the proposed method to attack a real-world face recognition system successfully.

关键词： Facial Recognition face recognition system Attackers model parameters

Efficient decision-based black-box adversarial attacks on face recognition

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Dong, Yinpeng Su, Hang Wu, Baoyuan Li, Zhifeng Liu, Wei Zhang, Tong Zhu, Jun Dept. of Comp. Sci. and Tech. BNRist Center State Key Lab for Intell. Tech. & Sys. Institute for AI THBI Lab Tsinghua University Beijing100084 China Tencent AI Lab Hong Kong University of Science and Technology

Face recognition has obtained remarkable progress in recent years due to the great improvement of deep convolutional neural networks (CNNs). However, deep CNNs are vulnerable to adversarial examples, which can cause fateful consequences in real-world face recognition applications with security-sensitive purposes. Adversarial attacks are widely studied as they can identify the vulnerability of the models before they are deployed. In this paper, we evaluate the robustness of state-of-the-art face recognition models in the decision-based black-box attack setting, where the attackers have no access to the model parameters and gradients, but can only acquire hard-label predictions by sending queries to the target model. This attack setting is more practical in real-world face recognition systems. To improve the efficiency of previous methods, we propose an evolutionary attack algorithm, which can model the local geometries of the search directions and reduce the dimension of the search space. Extensive experiments demonstrate the effectiveness of the proposed method that induces a minimum perturbation to an input face image with fewer queries. We also apply the proposed method to attack a real-world face recognition system successfully. Copyright © 2019, The Authors. All rights reserved.

关键词： Face recognition

Analyzing the Noise Robustness of Deep Neural Networks

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Mengchen Liu, Shixia Su, Hang Cao, Kelei Zhu, Jun School of Software TNList Lab State Key Lab for Intell. Tech. Sys. Tsinghua University Dept. of Comp. Sci.Tech. TNList Lab State Key Lab for Intell. Tech. Sys. CBICR Center Tsinghua University

Deep neural networks (DNNs) are vulnerable to maliciously generated adversarial examples. These examples are intentionally designed by making imperceptible perturbations and often mislead a DNN into making an incorrect prediction. This phenomenon means that there is significant risk in applying DNNs to safety-critical applications, such as driverless cars. To address this issue, we present a visual analytics approach to explain the primary cause of the wrong predictions introduced by adversarial examples. The key is to analyze the datapaths of the adversarial examples and compare them with those of the normal examples. A datapath is a group of critical neurons and their connections. To this end, we formulate the datapath extraction as a subset selection problem and approximately solve it based on back-propagation. A multi-level visualization consisting of a segmented DAG (layer level), an Euler diagram (feature map level), and a heat map (neuron level), has been designed to help experts investigate datapaths from the high-level layers to the detailed neuron activations. Two case studies are conducted that demonstrate the promise of our approach in support of explaining the working mechanism of adversarial examples. Copyright © 2018, The Authors. All rights reserved.

关键词： Deep neural networks

Towards robust detection of adversarial examples 18

学校读者我要写书评

暂无评论

Towards robust detection of adversarial examples

Proceedings of the 32nd International Conference on Neural Information Processing systems

作者： Tianyu Pang Chao Du Yinpeng Dong Jun Zhu Dept. of Comp. Sci. & Tech. State Key Lab for Intell. Tech. & Systems BNRist Center THBI Lab Tsinghua University Beijing China

Although the recent progress is substantial, deep learning methods can be vulnerable to the maliciously generated adversarial examples. In this paper, we present a novel training procedure and a thresholding test strategy, towards robust detection of adversarial examples. In training, we propose to minimize the reverse cross-entropy (RCE), which encourages a deep network to learn latent representations that better distinguish adversarial examples from normal ones. In testing, we propose to use a thresholding strategy as the detector to filter out adversarial examples for reliable predictions. Our method is simple to implement using standard algorithms, with little extra training cost compared to the common cross-entropy minimization. We apply our method to defend various attacking methods on the widely used MNIST and CIFAR-10 datasets, and achieve significant improvements on robust predictions under all the threat models in the adversarial setting.

关键词：

Understanding human behaviors in crowds by imitating the decision-making process

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Zou, Haosheng Su, Hang Song, Shihong Zhu, Jun Dept. of Comp. Sci. & Tech. State Key Lab of Intell. Tech. & Sys. TNList Lab CBICR Center Tsinghua University Beijing China

Crowd behavior understanding is crucial yet challenging across a wide range of applications, since crowd behavior is inherently determined by a sequential decision-making process based on various factors, such as the pedestrians' own destinations, interaction with nearby pedestrians and anticipation of upcoming events. In this paper, we propose a novel framework of Social-Aware Generative Adversarial Imitation Learning (SA-GAIL) to mimic the underlying decision-making process of pedestrians in crowds. Specifically, we infer the latent factors of human decision-making process in an unsupervised manner by extending the Generative Adversarial Imitation Learning framework to anticipate future paths of pedestrians. Different factors of human decision making are disentangled with mutual information maximization, with the process modeled by collision avoidance regularization and Social-Aware LSTMs. Experimental results demonstrate the potential of our framework in disentangling the latent decision-making factors of pedestrians and stronger abilities in predicting future trajectories. Copyright © 2018, The Authors. All rights reserved.

关键词： Decision making